Engineering Systems That Drive Sales (Replit Pro Series)

Video authored by the channel shown, not by IANIA; all rights belong to the creator. Transcript and summary generated automatically; the original-language content is preserved.

Original language: English

Original video: Engineering Systems That Drive Sales (Replit Pro Series) · Channel: Replit · 1:59:25 · Watch on YouTube

Topics: AI agents, Enterprise AI applications, Software engineering with AI, Markets and economy

Summary

Replit field engineers Ethan and Axel explain how they help enterprises move from AI prototypes to secure, scalable applications, and distinguish their role from that of sales engineers and forward deployed engineers. Through examples involving an Excel-based procurement tool and the migration of thousands of websites, they show how Replit can reduce costs and improve workflows. The discussion covers enterprise adoption, security and governance, requirements gathering, collaboration between technical and non-technical teams, and the emerging opportunities for AI builders and integrators. The hosts conclude that software is not dead: accessible AI tools create new opportunities for people who can identify real problems, build useful solutions, and connect them to users and distribution.

Key points

Field engineers work across multiple enterprise customers before and after a sale, helping with training, architecture, use cases, and adoption; forward deployed engineers typically focus deeply on one customer.
Enterprise AI adoption varies widely, and successful projects often begin with a knowledgeable internal champion who can identify a problem and explain its business value.
Hackathons and hands-on experimentation help employees discover practical applications for Replit and experience what software can automate.
Replit can help make legacy systems more useful by building modern applications on top of them, rather than replacing systems that still serve critical functions.
One enterprise used a Replit app to replace a complex Excel-based costing and bidding process, avoiding a SaaS purchase that would have cost $75,000 per year.
Another customer used a scraper and Replit’s MCP server to automate rebuilding thousands of websites, with the potential to reduce hosting costs by hundreds of thousands of dollars.
Enterprise projects require attention to governance, access controls, approved connectors, templates, and the context and skills provided to an agent before prompting.
AI enables business users, product managers, designers, and engineers to collaborate around working prototypes and code, shortening feedback cycles and reducing communication gaps.
The guests recommend curiosity, resilience, and iterative building: understand the business deeply, start with a focused problem, and add complexity as requirements become clearer.
The hosts argue that software opportunities are expanding, while distribution, community, and the ability to solve real problems are important sources of value.

Chapters

Start	Chapter	Summary
00:01	Introducing the pro series and enterprise value	The hosts frame the episode around moving beyond prototypes to applications that deliver business value, such as saving money or generating revenue.
06:06	What field engineers do	Ethan and Axel introduce their work helping enterprise customers understand, implement, and scale Replit.
12:35	How enterprises approach AI adoption	The guests describe uneven adoption and explain how teams identify valuable use cases and bring the right people into the process.
18:33	Company size, champions, and experimentation	The conversation compares the challenges facing large organizations with the flexibility of mid-sized companies and highlights internal champions.
24:22	Hackathons and an Excel-to-app case study	Ethan explains how hands-on Replit sessions create an aha moment and recounts replacing a fragile Excel workflow with an application.
30:14	Scaling enterprise solutions	The guests describe preparing a Replit template for the customer’s programming language and requirements, and discuss the value of replacing a costly SaaS tool.
36:34	Automating a large website rebuild	Ethan demonstrates a workflow for collecting website content and using Replit’s MCP server to help rebuild sites at scale.
42:26	Emerging AI builder roles	The guests discuss new roles for people who can build software solutions and emphasize learning through tinkering and practical work.
48:20	Is software dead?	The panel argues that AI makes more software problems approachable and that the industry is still early in discovering what can be built.
55:15	Custom software, distribution, and business objections	The discussion considers one-off software and changing business models before turning to enterprise concerns about security and governance.
1:02:40	Gathering requirements before prompting	Ethan describes listening to employees explain their problems and using that context to build a relevant application, sometimes in real time.
1:09:44	Prototyping and scoping work	The guests explain how voice feedback, prototypes, and iteration can help clarify requirements without delaying all building until planning is complete.
1:16:42	Cross-functional collaboration	A customer example shows business users and technical staff working together to build a secure interface around an existing SaaS product.
1:23:06	Prototypes improve communication	The panel explains how a working prototype gives engineers and stakeholders a shared reference and can accelerate feedback in large organizations.
1:30:34	Advice for aspiring field engineers	The guests recommend curiosity, collaboration, and a deep understanding of both the technology and the customer’s business.
1:38:00	Replit Live, community, and future programming	After the guests leave, the hosts discuss upcoming shows, improving production and clips, and the risks and opportunities of industry change.

Quotes

A field engineer is essentially a product / technical expert that works on the sales side. — 07:13

We’re trying to teach our customers to fish instead of do the fishing for them. — 10:58

They used the Replit application instead. — 29:12

Like this is the opportunity of a lifetime to solve problems. — 49:25

Speaking the same language. — 1:22:36

Concepts and entities

Replit, Ethan Torbinsen, Axel, field engineer, forward deployed engineer (FDE), enterprise, Replit Agent, Replit MCP server, Canvas, hackathon, ROI, C#, Excel, Blockbuster, Netflix

Full transcript (original language)

00:01 [music] [music] [music] [music] [music]

00:41 [music] [music] >> 5 seconds everyone. >> [music] >> Yo, what’s up? >> Everybody, welcome back to another episode of Replit Live. We are here in studio and we have an extra special episode for you all today. We’re going to talk about engineering systems that drive sales. This has been a little bit of a trend for us, talking about how do

01:13 you step in this professional role as a Replit builder and how do you start monetizing your business? We’re entering this era where building has become quite a bit easier generally and so we’re starting to see lots of people building all sorts of stuff, but who’s buying these things and how are you building a business around this stuff and how are you taking your Replit skills to the next level? So, we’ve got a couple guests that I’m really excited to introduce you today, but before we get into that, Manny, what’s going on, brother? >> Oh, I am pumped, dude. I’m a big fan of these two that are backstage. I’m so excited to introduce them to the community and to talk more about this theme around how do we get to value with

01:48 AI with Replit and a big part of the value is going to come from enterprise and who better to walk us through us through this than the folks that are doing it for Replit, and doing it every day. So, really pumped about this discussion. >> Yeah, and it’s actually a fun story. Uh, they’re brothers, and they’ve got an interesting background, and the way they came to be here at Replit, and everything is just, uh, cool to hear. And they’re also really, really smart. And they do this thing called field engineering. And so, we’re going to talk about what is field engineering, what is a forward deployed engineer. These are things that, like, you hear, you hear

02:18 these terms, and we talk a lot about integrating digital systems, or integrating these AI systems in existing, like, enterprises or businesses. And so, like, today we’re going to walk you through what that process looks like, you know, front to back. What happens when a company comes in and says, “Hey, we want to spend some money on AI, but we don’t really know how to, like, bring this in, or we’re not sure how to get the maximum out of this.” Well, this is where a field engineer, or a forward deployed engineer comes in. And we’re going to just kind of explain those things, and how that’s effective, and why you need to think about this. And it’s part of a pro series that we started, uh, last week. We started talking about how we’re going to build a viral growth engine. We

02:51 talked a little bit about how there’s this demo dilemma, and we’ll talk about it a little bit, I think, with the guys today, too. Where, when you’re showing off, like, from scratch demo, let’s say, from first prompt, that can always be exciting, because it’s like, “Hey, we delivered something.” But, I think the real challenge here, and what we’ve identified, is that showing people, like, zero to zero to one, is like, um, again, everyone can kind of build that. The hard part is, like, how do you take that one to, like, production? How do you take that prototype and put it in something real? Actually, Manny, the most recent episode we had recently was that prototype to production concept.

03:24 So, again, this whole concept of, like, um, you know, building production-grade software inside of Replit. And so, go watch the stream from last week. We talked about how we’re going to do that, how you can get involved, how we’re going to start showcasing some of the stuff we’re building, to really help you understand the types of things you can build with Replit, not just we’re building internally, but other, uh, um, departments inside of Replit, and other people outside of the space. So, Manny, any thoughts on just this whole, um, you know, pro series {slash} what we launched last week. >> Totally. The framing that’s coming to mind is how do we get you all to value

03:58 as soon as possible. And a prototype or something that you spin up really quickly in Replit, that’s interesting and it’s helpful if you’re learning something new, if you’re trying out a new feature, but that in itself is not valuable. What’s valuable is when somebody pays you for your application or when you get your team within a company up and running on that application and it’s streamlining their workflow, it’s saving them money or it’s generating new revenue. That’s the real value that we’re trying to get all of you to. So, I think that’s a helpful framing for me, um, Raymar, is like

04:29 again those prototypes cool, cool to show, cool to learn, but ultimately we want to get everybody to value and it’s really when they get that first user, when they earn that first dollar, when they get the team up and running on that app that they just built. >> Yeah, take that next level up. We talk about, you know, this whole era is not about like trying to simplify the technology. Uh, that will happen. The goal of this transition phase is to upskill your technical like upskill you technically so that you can step into this framework, right? Replit brings you up into this like new level where you can start playing uh, with developers and building real production

05:02 stuff with software and so like we want to kind of just tell you and help you harness all this. So, uh, before we dive fully into this, what’s up everyone in the comments? Hello summer camp. Epic in the house. I see KG in the mix. Uh, who else is here? Rushir, what’s up? Hello Brian. Kizzar in the house. Time to party. Good morning. >> tuning in from South Korea, I think. It’s like midnight there. >> What time is it even over there? Let us know in the comments where you’re watching from. Is that Ruth? He’s been I see in the house. Oh, hello, hello, hello. Uh, there’s actually Ruth, you’ll probably come up later today cuz we’ll

05:34 talk a little bit about how this all kind of started here. Who else is this? Oh, uh, oh, what can I think of his name here? Um, Gaddis. Gaddis is the one. Okay. >> [laughter] >> Nice. Let’s um, let’s bring the folks in, right? Let’s um Let’s just do this here. We’ll go to this shot. We’ll just Let’s Let’s not delay anything more. We’ll introduce the folks here. We’ve got Ethan and Axel in Well, Axel’s in studio with us. Ethan’s in the field. Ethan was supposed to be in studio with us, but he got

06:06 called out in the field. This is how it works here, right? Like things move super fast. You’re a field engineer, you got to get out in the field. And so, like we’ll talk a little bit about He’s actually out there doing things with the client right now. So, again, you know, we’ll just open it up, have questions. First thing we want to do is let’s introduce We’ll start with Ethan first. Ethan, I met you quite a bit ago. And actually, that’s why I said Ruth would come back up cuz there’s Ruth saying Ethan. We met doing these like build hour things like years not years ago, but like what maybe a year ago, year and a half ago, something like that. I don’t know. We just like little rinky-dink meetups. And we were all just like kind

06:39 of passionate about Replit. And you and you were in there, Ruth was in there. There’s a a handful of other folks. Um >> And this is before you guys joined Replit? >> Long before joining Replit. Yeah. Yeah, yeah, yeah. So, tell me a little bit about like What is a field engineer? What do you do? Who are you? Just give me a little rundown of who is Ethan Torbinsen and what do you do here at Replit? >> Yeah, well, I’m super excited. It does feel a bit full circle just because obviously we we did those kind of build hour buildathon sessions. I think it was a year ago, actually, which is pretty crazy to think about.

07:13 Um But yeah, hey everybody. I’m Ethan. I’m a field engineer here at Replit. A field engineer is essentially a product / technical expert that works on the sales side. So, we do a lot of things pre-sales and post-sales where we’re supporting um the sales process for large enterprises. So, we do a lot of trainings, we do a lot of talking through architecture, we do a lot of um kind [clears throat] of supporting large enterprises as they try to scale

07:45 their Replit usage. So, there’s going to be a lot of fun stuff we’re going to talk about today. Prepared some stuff, but still super cool to to be with Raimar and and see all of you guys. >> It’s one of those things where we talk a lot about how you don’t know how things converge when you build community or where when you start connecting. So, you know, you never know where like if you just start a hangout, you start getting people together, you start doing things. You know, maybe somebody from that group’s going to go get a job at like your favorite tech startup and maybe like they’re going to bring their brother along with them and I don’t

08:16 know, right? Maybe you’ll be doing cool live streams. I don’t know what it is, but like we talk about all the time. Go start building your little community and hopefully again we’ll start exposing you to like some of these up skills, but Alec Axel, I’m sorry. Tell us a little bit about you know, you came along as a as as a you know, a field engineer also with Ethan, but tell me a little bit about your background cuz we did not meet. You were not in the build hours. >> Yes, nice to meet everybody. My name is Axel. I’m a field engineer joining my brother here. Started about 7 months ago at Replit and have been working on you know, working

08:49 with our largest enterprise customers to help them implement Replit and help them use Replit in their actual work. You know, our goal is to basically understand their technical requirements and their technical guard rails that they need set up and then apply those and let them kind of build within those while using Replit. So, yeah, excited to show some stuff off and talk a little bit more about what we do. >> Okay, great. I think the first thing we need to do is help people understand like um the um like you just said what a field engineer

09:21 is, right? Going out and helping customers, clients, etc. But I think people think like forward deployed engineers do the same thing, too. So, maybe you guys can help me understand the difference between a forward deployed engineer, a field engineer. This is kind of an emerging thing, right? Field engineers have been a thing for a a but we’re seeing them really kind of step up, especially as like lots of people are looking to integrate this AI stuff in their business and they don’t really know how to start. So, like, you know, that’s where you all come in. So, tell me a little bit and uh I don’t know who wants to start here, but >> Yeah. >> Uh actually, yeah, go ahead. Um what’s the difference between the two roles here?

09:51 >> Yeah, you hear a lot of terms uh around this space. So, there’s like things like solutions engineers, sales engineers, and then field engineers, forward deployed engineers. All of these types of things have slightly different focuses of their role. And field engineers actually uh something that is a little bit more specific to Replit and kind of uh is is our take uh on the philosophy of using AI in the future and like being very agentic and kind of spreading the aperture of things that you can do. Um so, typically a sales engineer will be working on like pre-sales. They’ll talk about um

10:26 you know, what the platform can do. They’ll give demos, things like that. And then sometimes they’ll not even be on the call post-sales after somebody’s closed the deal. A forward deployed engineer uh will focus on a specific customer. So, they’ll kind of get really deep in one customer and they’ll just work with them. They’re really focused on their use cases. We as field engineers kind of have a broader um broader things that we work on. So, we work with a bunch of different customers. We work pre-sales and post-sales. We’ll sometimes work on specific use cases, but a lot of times we’re trying to teach our customers to

10:58 fish instead of uh do the fishing for them. Though, I will say sometimes it’s great to just do the fishing for them. Solve their problems right away rather than like teaching them how to solve their own problems. So, that’s what we’re focused on. We’re focused on uh kind of helping a bunch of customers get really empowered with Replit and get really into the weeds and understand how it works and then build out uh from there. >> So, you’re kind of doing uh at the enterprise level, I think really what we’re trying to do here. And I think that maybe that maybe that’s a shift, man. Maybe we have to shift this like Replit Live. Like, we’re just uh virtual field engineers. Right. [laughter] Coming in your inbox because, you know,

11:29 most of y’all are not enterprise clients or like maybe have access to an Axel or an Ethan all the time. And I think that’s where some of these live streams come in. So, yeah, maybe we’re going to start thinking about how do we fill that gap a little bit? And I think that’s part of what we’re seeing this pro series. Ethan, any um any gaps you want to fill between what Axel just said there about difference between field engineer or forward deployed engineer or anything else, you know, to button up that little segment? >> I I I think you nailed it. I think really the a forward deployed engineer goes very deep on a specific customer, and then a field engineer has a very large breadth. So, we work across many,

12:03 many different customers. Um we really try to um do do as the best we can to very quickly unblock them. Um so, that’s kind of the the way to to think about it. >> Can can can you help us um appreciate the context a little bit more, Ethan and Axel? So, like I feel like enterprises have really gone strong on the AI front the last 12 months. Maybe like walk us through what you’ve been seeing, the uptick in interest on the enterprise side. And then maybe talk a little bit about why you need this type of hand-holding cuz, you know, if people

12:35 think of Replit as a consumer product, like well, why can’t people just do it themselves? It’s different at enterprise because of the integration part and getting teams up upskilled and leveled. So, maybe talk a little bit about why this function is so important particularly on the enterprise motion. >> Yeah, I will say one thing is um enterprises are adopting AI, but it’s very unevenly distributed, right? There are some enterprises that are super in on it, that are really building out crazy cool things and completely restructuring their business and redoing how they work. And there are a bunch that are not doing that, right? There’s a lot of momentum to how uh

13:09 previous processes work. There’s a lot of momentum to how everyone’s done things for years and years. And so, it can be a struggle to actually implement and get things uh get the right tools in the right people’s hands and things like that. So, our focus is to help like if if a if a company wants to kind of transform their business, our focus is helping them figure out how to do that. Sometimes it’s focusing on specific use cases or specific departments that can, you know, have the maturity or have the understanding of how important AI is, and then implementing it there. Um Yeah, it I will say it varies a lot by

13:41 enterprise with how how mature for things are. >> Well, it’s evolving so fast, right? Like I mean, I have to imagine lots of enterprises know they need to be doing this stuff, just like lots of businesses need know they like everyone’s hearing AI, you got to bring it into the business. We saw this explosion of like token maxing for a while now, but I think now everyone’s in like the okay, but what does this actually do for me phase, you know? And so, maybe you can um help us think like how are enterprises deciding what to build? How do they select? Like how do you help

14:12 them think through this? Like take me through okay, you’re you’re meeting somebody for the first time. How are you thinking through this? What questions are you asking? What are they like bringing to the table? Like let’s kind of walk through this this whole process if we could from from top to bottom cuz I think it’d be helpful for people to just, you know, think through that. >> Yeah, and it it’s actually interesting. I’ve heard I’ve heard you all talk about this a lot as well on previous like live streams and community events. Um a lot of times it starts with like a subject matter expert that has a huge problem that they want to solve and that they

14:44 just know they know there’s a solution to it and they can’t do it themselves. Uh and so we we have a a huge growth of Replit enterprises. Somebody trying Replit out, realizing that they can use it to solve their problem, and then coming to us and being like, “Okay, I solved my problem, but I need to be able to like do it very safely and govern it and scale it.” Um so, yeah, that’s that’s like this the kind of quintessential thing that happens, right? >> That’s kind of like the dilemma in the moment right now, right? It’s like easy to build that prototype, but this is what we’re talking about where it’s like, “Okay, it’s fancy, it’s fun, and

15:15 it works in your local machine. You’ve got it like set up. You got this cool little like dashboard built, whatever, whatever, but like now it needs like real data.” Or now now it needs to touch this enterprise system that’s locked down. Like those are the those are the things, right? Like so how how are you how are you seeing how are you thinking about that? How does that like Yeah, how does that play out? >> Yeah. Um So the the first thing is understanding like the importance of the use case and then how their business works and how much value and ROI that use case can be. And then you’re to your point, it’s like starting to connect to those internal systems. You might have a database that has like a ton of internal

15:47 data and you don’t want somebody to be able to just connect to all of it uh because they might be pulling in things that they either aren’t allowed to see uh or you know, just shouldn’t be applied to their use case. So then we’ll help with like governance controls around, you know, you might have like role-based access controls. Only certain people can access certain things in their database. Or even row-level security like I can access this table of data, but I’m only allowed to see a couple of rows in it and then my co-worker can see all the rows and you know, have vary that. So it’s kind of use case by use case, I will say. Maybe Ethan, you have thoughts on on how we’re

16:20 starting to to see this go. >> Yeah, I think it’s usually um people will I I and I see this a lot in large enterprises where there’ll be somebody who um who often has been playing with some of these tools and then they discover Replit and very quickly they’re like, “Oh my gosh, I could solve my problems with this.” Um and it’s especially helpful if they’re a little bit more senior in the company because then it gets a little bit more momentum. Uh but as soon as you get that that like that um that expert in their field who

16:56 sees a problem, really understands their problem, and can articulate that ROI, then we are able to kind of start getting momentum at the enterprise level where we’re able to loop in the the people that are actually going to be purchasing the software um and then we can start talking about governance and scaling this um so that more people can use can use the softwares. Um the I think the key aspect of of Replit is it just and we all know this um but it just gives so much power

17:30 to the people who understand their problems. Um it’s like you you used to have um an experience where there people with um who who kind of had this domain expertise were quite limited in what they were able to do with software but now it’s once you’re you’re opening that up it really transforms a business and it’s just it’s getting in the right rooms with the right people um like the security teams and showing them how we have uh we’ve built a platform that allows you to control um

18:03 and kind of have some guardrails on the projects and and and data that people have access to. >> Mhm. We talked a little bit about AI adoption not being evenly distributed. Can you two talk a little bit about the sectors in which you’re seeing the most traction for tool like Replit and also within those sectors what departments do we see like most excited about Replit? >> That’s a that’s a really tough one. It’s very it varies so much. Like it’s funny we’ll sometimes >> industries

18:33 >> Yeah so some other uh technology companies um will have their like team split out by industry vertical. We don’t do that. The the reason we don’t do that at least so far is because we have are working with just such a variety of people it’s kind of crazy right? Um but I will say that I have been seeing a lot of you know mid mid-sized companies that are able to really transform. You know with with a huge company there’s so much momentum in how they’re doing things. It’s really hard for them to actually like change a whole bunch of

19:04 processes. Like I was I was uh there’s a story that I that I was listening into from a high-ranking person at a large financial institution and they were they were talking about they were talking with some of their other uh people that were working at the company about how many things AI could change about their company. And they’re going back and forth, and they had all these ideas about all these cool things that they could do. And then one of the people chimed in and said, “Yeah, this all sounds great. There’s no way we can do this. Our company is set up. Our processes are set up for us to not be able to move the company in this direction, to kind of

19:38 make that huge pivot.” And that’s just how it is for certain enterprises, right? You you’ve kind of set your company up on a certain trajectory, and it’s hard for very, very large companies to do that. Whereas with a mid mid-size company, or even certain large companies that have a lot of uh kind of ambition and resilience to change, you can pivot your company and change so that you’re able to adopt AI at the scale that you need to. >> Yeah, I think that’s going to be an interesting thing to watch. We’re about to see a lot of Blockbusters get Netflixed, you know, uh because of

20:08 exactly what you’re saying. And for a long time, right, having these established processes and procedures, like this is how you build a business, right? Like you create systems, you put people into those systems, you have them operate as some kind of like knowledge worker assembly line, right? In some sense, where like people do different things, you kind of isolate responsibilities. And now we’re entering this phase where like uh you know, the generalist is winning. Somebody who can kind of see across departments brings a superpower, cuz like you said, they have like some unique insights about gaps that, you know, they can fill with these tools.

20:40 And you know, like I don’t think it like the size of the business matters. Uh I think it’s more about that mindset of like bringing in the shift, right? Understanding that like you might have to change some things, or at least empowering You talked about like um you know, champions a little bit before, getting like a an advocate in in the game. And so like having someone who helps uh bring this to the table. But yeah, I don’t know. Um >> Yeah, and one thing on that, you you mentioned, you know, subject matter experts doing things and um you know, people people who like know a lot about a particular probably doing

21:13 things. That’s a huge That’s a huge part of who our kind of like Replit champions are. So, a Replit champion is somebody in the company that’s really advocating for us. But, another thing we actually see, too, is not just subject matter experts, but people who have a lot of ideas um and maybe they aren’t like super deep on a particular topic, but they have a lot of ideas of things to build and a lot of uh lot of skills at interacting with those subject matter experts, they can be super successful as well. So, it’s not just people who have, you know, been in an industry for a super long time or know the area very

21:44 well. It’s just people who have a lot of like ambition and drive to solve problems and have a lot of ideas about how to solve them. And then they can just go go go go go and they’re very Another thing is you have to be very resilient, right? When you’re when you’re building these types of uh projects, you have to be able to kind of roll with the punches as either technical things break or you learn that requirements change. Um yeah, it’s it’s super cool to see the different types of people. But, yeah, subject matter experts, super great. People with ideas, people who are more generalist, super successful. >> This is one of the interesting things

22:15 for me when I hear about some of these processes and when you all go do things like hackathons, Ethan, maybe you can talk about, you know, like what you’re I don’t know if this is what you’re doing out there or if you you know, you have experience with this type of stuff, but like exactly what you’re saying, Axel, is like tapping into the knowledge of different people inside the org, right? For so long, the only people who could like work on these systems or make modifications to the processes were people who were technical who could kind of change the underlying code, the infrastructure of how these systems operate. And now you can go into an enterprise, you can say, “Hey, we’re going to give you all a Replit

22:46 workspace.” Or you know, like and we’re we want we want to like what all are your ideas? And maybe again, those things don’t get pulled into production, but you start getting the ideas from the HR team or the sales team or the whatever different team that may have never surfaced before, right? So, even if it’s not like, “Hey, we’re going to let everyone build and push to production on our like, you know, pro-level product.” You can still prototype at a deeper level. You can still get buy-in. You can still let other people kind of have access to some of these tools internally. So, Ethan, I might um you know, throw this to you as people like

23:18 you know, or or especially inside of enterprises, like how how are you seeing them kind of tap into more ideas, uh empower folks that are non-traditionally like in the decision-making role or kind of impact the product role. Like, how how how are people handling that? >> Yeah, I and and like we’ve said, it definitely varies company to company. Um you you see sometimes there are like these top-down mandates to use more AI. Um you also see people kind of um kind of a grassroots approach where it’s

23:49 like they’ve been given tools and now they’re able to kind of solve their own particular problems. Um to kind of take a step back, I think one of the reasons why we focused on uh ac- actually, I guess we haven’t really talked about this as much, but uh a big favorite of ours in the sales process with enterprises is to hold some sort of hackathon, um where you get a bunch of people in a room um or online um and you basically give a demonstration of how to use Replit and then you give

24:22 them time to go and explore. And so, um we’ve found that that has been so helpful um in in terms of getting people excited about the solutions that they’re they’re building, the the problems that they can solve because uh I still remember um when I first had like wrote my when I wrote my first Excel macro, I just thought it was the coolest thing. I I was in middle school and I was in like some advisory class and I just

24:56 remember like doing some sort of macro that like move numbers around and I my mind was blown that computers could do that. And and I think once that that like once you kind of understand the things that you can automate with software it it just completely ship flips the switch in your brain of what’s possible. And so the reason why we we focus a lot on hackathons is to try and get that like aha moment where it’s like oh my gosh

25:29 I can actually solve the terrible and annoying Excel sheet system that we have and make a an app for this where instead of having like a bunch of Excel uh kind of cells that it can just have a form or connect to existing data and it can save me hours a day. um in in my time and we see that like we see this a lot in in these

26:01 um these office hours as well where after after we’ve kind of been implementing Raplet at at an enterprise people will start saying oh my gosh my job is fun now because I can actually just I I have the ability to do the like actually go out and test my ideas and and solve problems and not feel so limited. >> Can you totally I I remember yeah my first macro it’s so for everybody like in Excel like you can go in and drop in

26:33 work in cells but macros allow you to codify something and add like automation stuff like that and usually most people is foreign to code and so that first time we have like a macro you push a button and runs all the things that you did manually you’re like oh my gosh this is amazing I’m never going back but it’s hard to kind of write those and maintain those but now we have AI to help us with that. Um can you maybe walk us through a an example a case study like maybe walk us through like the full flow like how this prospect came in or this company came in what were some of the initial hesitations and hurdles that you had to overcome and like ultimately where did

27:06 where did they land like what what value were we able to deliver to them? >> Yeah, so good example actually staying on the example of Excel and macros and things like that. So we had one company who was a it was like a large-scale enterprise that was working in the aviation and transportation space. So they’d sell like technology and hardware to other companies and they’re like a very large company and one of their data analysts discovered replit I think actually was

27:37 through just like a YouTube video that we had posted. I don’t remember the exact YouTube video. I wish I would have wish I would have remembered that so I could give it credit. But he discovered replit through a YouTube video that we posted and started playing around with it and realized that he could take some of the the kind of modeling that they’d done in Excel and move it to a much more easy to consume format. So they had like they had this gigantic Excel document and sheet and whatever that had a bunch of equations that would always break cuz you know every single cell and in the Excel is like an equation and you have to go through every single one of them

28:10 and there was it was only him that understood this Excel. And so all the all and this this Excel by the way was used for like costing and bidding on these large-scale uh these very large-scale um you know the what what would you call them? >> Procurement systems. >> Yeah, procurement systems. Yeah, exactly. So doing a huge huge procurements and so this was like this was being the basis of a ton of money that they were spending and only one person understood it him and and so all the people who were kind of

28:40 business experts didn’t understand so they’d like tell them a bunch of assumptions to put in and he’d like play around and make mess around with some of the things and then they would like go back and forth and back and forth, and nobody had a really clear understanding of what was going on. So, what he did was he took that, put it he just basically uploaded it to Replit, went back and forth with it, and then created a Replit application that allows them to do all that bidding and costing in one super simple place with a really easy to understand interface, so that he could send it to his other business experts. And then, after they did that, they ended up

29:12 being able to make much more informed better bids. They also were thinking about actually buying like a SaaS tool for it that would have costed them about 75k a year, and then they just didn’t have to buy that. They used the Replit application instead. And then, at that point, they were like, “Okay, this is amazing. We need to scale this across the enterprise.” So, then they came Yeah, so they came so [laughter] they came >> Pay us more money. >> They came to us and they were like, “Let’s Let’s get this going.” And so, what we did is we went in, figured out how their systems fit together, talked we we talked to them about how to like

29:44 integrate some of their databases and things like that. They also have like for their production systems, they also were written in a specific programming language, C#. And so, Replit can support a bunch of different programming languages. So, what we did is we made a template for them, you know, cuz Replit Agent will pick the programming language for you if you don’t ask. But, if you want to write in a particular programming language or have Agent write in a particular programming language, it can do that. So, we set up a template for them so that their users who wanted to bring things to production eventually would kind of out of the box

30:14 already have Agent writing in their exact programming language with some of the kind of specifications that they have. So, they have like a security system that they wanted to connect to. And in this template, we included all this stuff in it, so that they could kind of just go right from like prompt to their engineers wanting to put it into their production system, and they’re like, “Oh, this already looks like what I would push. You know, this already looks like it has the things that I would already want to integrate into my my system.” And so, they could they made that process much smoother for them at that point. >> You mentioned there were we

30:45 What were the What What’s What was the alternative? So, like at the end of the the engagement, what did you show them? Like, here’s how much we saved you and like in terms of like SaaS options or like here’s how much we generated in terms of revenue. Like, what would like the final number or like estimate? >> Yeah, so for that one use case, it was 75k that they saved for the the SaaS thing. But, that was one use case, right? And that was one dude. We have since scaled to I don’t actually I think it was about I think it’s about 75 people at the company and then they’re just looking to kind of increase and increase and increase. And every single one of those people has use cases and

31:15 ideas like that or similar to that. So, that’s like the easiest ROI. But, they’ve done other super cool use cases where they’re like working on it to make their actual like creation of the technology and selling of the technology more efficient to make the customized sales process for them more efficient and better. Um Yeah, there’s a lot of >> a great example. >> How many places do you go into and it’s just that? It’s an Excel spreadsheet running the whole shebang. >> Like, it seems like, you know, everyone thinks like, “Oh, everyone’s like building their own software and we’re like just, you know, we talk about how

31:46 you’re we’re early into this.” So, yeah, Ethan, Axel, like what do you see and like how like are these systems fairly sophisticated people have or a lot of people just kind of running these spreadsheets and Google Docs and they kind of got these like systems that like, you know, executive age running something over here and sales team over here is running their own little thing and they’ve got these little silos and then what where the real value comes in and this is where I keep trying to like focus people is like this is the thing that nobody sees on the outside is these little systems that are being built. Last week we talked about like the real interesting story is like what’s happening here inside of Replit as we all kind of build this AI first company

32:18 and we all start using these tools internally and a not a lot of that breaks out of the surface, right? So, a lot of people are left looking like, “Where’s the real value of AI?” And we see it and so we’re like, “Oh my god, look at what you can get out of it.” But, the average person doesn’t. And so, I you know, like these are the types of systems that people are building that are never going to like showcase because like an enterprise is just never going to be like, “Hey, look at all this money we’re creating with these systems that we’ve optimized, right? Like this is going to go to the bottom line. Shareholder value’s been created. Let’s [laughter] go.” You know? So, I I I I wonder like how many of these systems that you go into are basically just

32:49 that? Like they’re just spreadsheets, they’re these disparate systems, and what you all are building is like glue to bring them together and help kind of optimize the connectivity of those or >> So common. So common. It’s It’s actually so interesting. Um and it’s one of the things that I love about Replicon and also just modern technology. So, you go into an enterprise and they have systems that were created uh before, you know, 20 years ago, 30 years ago. Um and they’re they’re very old, but they function, right? And so, some of them some of them they need to replace uh because they’re

33:21 just not scaling right. But some of them they actually just can’t replace even if they’re super duper old. So, what they’ll do then at that point is, to your point about glue, they’ll make modern applications on top of these really old systems, right? If something You know, there’s that saying, “If it if it ain’t broke, don’t fix it.” Some of these systems are just so critical that you can’t really rip them out and place replace them. Some of them you do want to, I will say. Uh but for those ones that you can’t do that, building things on top of them is so helpful, right? You have this old legacy system, you build

33:52 something that’s like very modern and useful. And I will say to me, that’s like one of the most beautiful things I see is like you have this system that was built 20 years ago or more, and then it’s like existing with this thing that is just as cutting edge as you can possibly be, and they kind of coexist in the technology space. I just absolutely love when those types of things happen and those types of use cases come through. It’s just so cool. >> This is where I think the enterprise software is going to be. It’s going to become the rails. We talk about this, right? And the new invention is like connecting all these dots. Ethan, I saw

34:23 you shared your screen. What thoughts do you have here as we’re kind of chatting about some of this? I know you’re probably itching to dive in here. >> Yeah, before you before you share that um for the live stream, I’ll just give a little bit of context. There is uh an enterprise customer that we were working with that was a large it was a a marketing kind of boutique firm that had a very kind of nail down niche and and they were they were hosting thousands of websites in a particular

34:55 hosting platform. It was like basically archaic. They they had websites for these for all of their companies and it was kind of like one of their product offerings, but it was so bad like basically everything was broken. It was just a it was a very simple front end, but it was all like 1990s looking stuff and um and they were trying to figure out how to to revamp that, right? Because their their particular platform that they were

35:27 using was not built for managing thousands of websites. And how do you approach like redoing those? And so we so we were talking with them and they were basically like, “Look, this is how much it costs to host each website and it basically came out that they were spending over $500,000 a year in hosting costs um just just to have those bad websites.”

36:03 And so we we talked through the like the value of switching hosting over to Replit cuz Replit has very cheap hosting. And and we determined that like if we got these websites over into Replit we would be able to save them literally overnight hundreds of thousands of dollars. Now they were trying to figure out a way to to programmatically do this. And lo and behold, Replit has an MCP server.

36:34 Which is awesome. Um so, basically what what they were doing is um they were trying like on their own they were trying to rebuild each website in Replit and it was kind of it was taking a while, it was costing them some time. And then um the account manager brought me in and we talked through their their problem and really what what we just needed to do was um automate that that whole process. So, we built

37:05 we built together um in Replit a scraper where they would just upload a an Excel doc of all of their websites. Um it did it an amazing job of compiling all the content, all of the um the images, all of the links, and then packaged it in a way that Replit could leverage to build it using the Replit MCP server. Um and I could just if you want to show the screen, this isn’t the

37:36 actual thing. Um I rebuilt it as a demo for this um just so you could kind of see what was going on. But you think about this and this is thousands of websites, right? Um so, they need to to have visibility into kind of where things are at in this. So, basically they uploaded uh they would import um with their site list here. And then um once it once it got kicked off, it would show up um as a obviously demo,

38:07 but whatever. Um it would show up here as kind of like in progress, what what what sites are being built. Um then it would have a review process. Um you know, a second. Maybe it timed out. Um but essentially what this allowed them to do was scale their uh their website redesign process that um would have just been impossible to do without AI. And so um within it was crazy. Within 24 hours we

38:42 had this and they started rebuilding them and um they were basically completely shocked and it is they they actually kind of had an identity crisis and trying to figure out should we do this full-time? Like should this Is this the thing we pivot to instead of being a marketing company? Should we just do rebuilds of websites? So >> It’s so funny. It’s so funny. Ethan uh you know Ethan was obviously involved in this and he can’t he can’t talk himself up too much. But like in 24 hours they like hopped on a call and then

39:14 Ethan built out this use case for them and talked to them about the MCP and literally completely revolutionized their business just because of like the knowledge and new things that they could do. What did they call it now? They they wanted to sell it. What did What did they call the the system, Ethan? >> Uh I think they called it the Torby 3000. >> What is [laughter] that? >> The Torby 3000. >> They they started calling it the Torby 3000 just because Ethan like had done so much work for them. Like like the the ability to like rebuild all those

39:44 websites at like a fraction of the cost, be so much higher quality, and then also save hundreds So you saving of thousands of dollars on hosting costs and then you’re also making them much higher quality marketing websites. And they’re just like woah. >> Can I pull out a theme from these two stories cuz somebody asked earlier like are we prompting for them? And no, it seems like these folks already know about Replit, they’re getting their hands dirty, they’re try they’re they’re testing it, they have a prototype, but then they need to think about A, how to integrate it into their system and B,

40:16 how to scale it. that’s when they’re reaching out to Replit and the field engineer. So, I wonder if there’s like some not like some reference there for our folks in the audience that are also trying to do this where like it’s not like you want to like have that first call and be like, “Here’s an introduction to Replit. Here, let me show you how to prompt this thing.” It’s almost like you want like a whole bunch of folks to try Replit, and then some percentage of the whole folks are going to have that wild moment on their own, and then they’re going to say like, “Okay, I have this prototype. Now, like like I need to bring you in to help me think about how do I roll this out to the entire enterprise?”

40:49 >> I want to tighten one thing on top of that, Manny, because this is we talk about folks like like right now Ethan and Axel are talking about how they do their job, and they are obviously like paid field engineers here at Replit, but there are opportunities for you all out there to become these integrators, to become the people that bring these systems into these enterprises, that show them how to do this, right? Like it doesn’t just have to be coming from Replit. Obviously, like when they’re considering Replit and when they’re in a sales relationship, it behooves us to have somebody that can go support those types of roles, and make sure that like the company has what they need to go and

41:19 think through this, but you all out there, as you’re upskilling and as you’re like learning with Replit, if you’re building an agency, if you’re a freelancer, like there are so many businesses that just need help pulling this stuff into their business that you can become their AI champion, right? Like you don’t have to be an employee of theirs. You can be the person that introduces it. You can be the person that like, you know, we’ve talked before about like taking some restaurant and rebuilding it for them as a prospect call, and sending them the new demo, and be like, “Hey, I built this for you. Do you want this? Like you can have it for free, but just hire me to like keep working with you.” You know, and so like there’s all sorts of ways to break in.

41:52 And so like part of what we’re doing here is trying to expose you to like this is how we’re doing it internally, but also to get you to lock in and think like how can I apply this back to my world, right? Like because it’s easy to see this at a high level and think like, “Okay, yeah, but that’s for them with big clients and whatever, but how do I do this?” So, uh Ethan, go ahead. >> And Yeah, and and I would say that seeing emerging roles because of this technology. Like we see a lot of companies that like the reality is that a lot of like a lot of us here are kind of more early adopters, tinkerers. We love to build

42:26 things. Um and some companies that’s just not the case. A lot of people just like to do their 9-5 and they kind of do their work and that’s that’s totally fine. Um that being said, there are a bunch of roles that I’ve I’ve started interacting with that are essentially AI builders. Like that’s what their role is. I actually in my previous role, that’s what I was basically. I was essentially a solutions engineer and that was just a fancy an AI

42:56 solutions engineer, which was just a fancy way of saying um the team would come to would come up with with a problem or like talk about some sort of problem that could be solved and then we would create web applications to solve those problems. And I would be the primary builder on that, but would work very closely with those teams to help them get set up with that. And the reality is that um like those these roles

43:28 they’re becoming more common because it’s like some people are just that’s their more of their tinkering mindset and and they can solve problems like that. And then some people are um not as oriented towards that and that’s okay. It’s just um so I I think there’s a lot of opportunity and I think an emerging job they’re emerging jobs with that in there. Um >> And and one thing you’ve been saying is if I go if it you don’t mind I go on a slight tangent here. >> it. >> Um you know, what one of the things that got us started in Replit was

44:02 actually that tinkering mentality that that kind of we have ideas, we want to build them out. So, we actually started on Replit before Agent was released. Before like Agent 1 was going before there was even this concept of vibe coding because we wanted to work on projects together on the weekends and like do coding in the same place and Replit has the online IDE where you could type, you know, you have one person in one file typing a functionality in code and then another person in that same file typing things out. So, Ethan and I actually started tinkering on Replit like years and years

44:32 ago. I actually still have the $7 Replit subscription that gives me $20 in credits because I’ve been grandfathered in over the years, which I’m very excited and lucky for. >> By the way, when you said that the other day and I’m John was like, “What?” >> [laughter] >> They keep They keep threatening to take it away. It’s like, “Come on.” >> Yeah, take it from my cold, dead hands kind of thing. >> Yeah, [laughter] I’ve been You know, we we talked to the the engineers who like kind of own that and they’re like, “Gosh, can’t believe there’s people like you.” >> Legacy plan. >> Yeah, let’s go. Um but that’s how we got started, right?

45:03 That That was building stuff on the weekends, tinkering, doing side projects. It’s like the best way to learn is to just make stuff, right? And that’s one of the things that’s so addicting about AI is that you’re just building things and if you can build things that are useful, it helps you learn so fast and it’s just very fun to do. It’s kind of the best of both worlds. It’s like very fun and you learn a lot. >> I want to call out a question here from Scott. Do we offer official training on how to get started with with Replit ASAP? You’re in it. I think you’re in it. This This is it. You’re you’re watching from from from the front lines

45:37 in terms of what they’re doing how they work. Ethan and Axel just mentioned like they were tinkerers, they’re trying things out. Axel Ethan was talking about how he was building internally within his own company organization. I think that is the best training. You can come here, you can get get a sense of what we’re doing and get inspiration and you can show up to Friday and figure out what’s new, but ultimately I think the education is going to come by you tinkering like Axel did or like rolling it out internally within your own organization like Ethan’s been doing. >> Can you transfer the legacy plan over to your children? >> [laughter]

46:07 >> I got to try. I’ve got to try. >> That’s amazing. >> Yeah. >> [laughter] >> Get something signed by Amjad. >> That’s great. >> At a party or something. >> I Yeah, um you know, this it’s the same stuff, right? Like I know people are like maybe you were expecting some, you know, like breaking news here as we were going to like tell you about solving problems and bring in some enterprise folks and like think that we were going to talk about anything different than like the same stuff we kind of talk about on a regular basis, right? This is part of what we’re trying to tell people is that everyone’s

46:40 kind of facing the same thing. We’re very early to this. Like there’s lots of people trying to figure out how do you plug this in, right? All the way up from organizations like Replit who are kind of pioneering this technology, all the way to the ground level integrators. And so, uh lots of folks talk and maybe we can dive into this a little bit. Uh there was a viral post this weekend uh from Levels. io that was like software is dead, you know? Uh and I saw that and I was like no like you’re crazy. Like the we haven’t even seen the beginning of software, right? Like imagine thinking there was going to be less software after the printing press or or less text, right? After the

47:14 printing press came out. Now, does that mean that we’re going to see the proliferation of like more enterprise grade Salesforce type stuff? I don’t know, right? Like maybe that type of thing is going to be harder to build. But these little like the little projects, the the things that people are going to build and then again the ability to like monetize that at whatever level or find optimizations inside of your business. So, it’s like all these different sides of the same coin that we’re trying to expose folks to here. So, I don’t know if we you know, like do we derail a little bit and just think about like, you know, what what do you think? Is software dead?

47:46 Obviously like these enterprises are thinking about what can I replace? But then they’re building new stuff then we’re seeing byte code and all these different people are like entering the creator sphere. So, like let’s maybe riff a little bit on this idea of like where this whole industry is headed, how people can prepare for this. Like, you know, you know, like is software dead, I guess, is the question, you know? >> Yeah. And And before we get there, one thing that makes me think of is and I’m I’m curious for a pulse check on the room. Do you all feel like you are using AI to the like uh maximum capability it has? Like, do you feel like you’ve

48:20 actually completely integrated it into your uh day-to-day workflow? And do you feel like you actually are like fully utilizing it to to the maximum degree? >> Let us know in the chat. Have you fully integrated this? >> Because Cuz I will say I work I work at Replit. I do this all day and I feel like I haven’t taken advantage of like so many things. I have like a hundred ideas in my head of things that I need to build. You know, I’ve I’ve automated huge parts of my job um and done done a lot of work on like creating the skills and the kind of brain that powers the Axel agents that

48:51 go off and do work. But oh my goodness, there is so much to do uh and you know, to your to your point about like is is is software dead? Like there’s just so much software to still make. Like, we have so Yeah. Yeah. We’re Yeah. Barely scratching the surface. Yeah. >> I think this is very comparable to the printing press in the way that it’s like we now are unlocking the ability to solve problems that we just could never even attempt before. So like the like before the constraint was obviously it’s like you need software engineers,

49:25 you need a lot of resources to be able to to build every single kind of idea that you have. Um but now it’s just people are able to attempt more problems and that means the number of problems that can be attempted is like a thousand times, maybe a million times larger. And so I think if anything it’s like software is just getting started. Like this is this is like the opportunity of a lifetime to solve problems. Like I just

49:58 [clears throat] if you have if you have high agency, you can just do you can just do things, which is awesome. >> This cracks me up here and Nadi is saying I come here for the community stay for the printing press analogies. We make this analogy all the time and I’ll make it again for those who haven’t heard it because the printing press was an exciting piece of technology, but it was only possible because of there was like advancements in underlying technology. We talk about advancements in the printing the ink, eye glasses, paper at scale, like all these different things kind of were coming together and we’re in that moment right now with the

50:29 web. We’re like the browser is super powerful. The AI at at a core level is super powerful. You have like insider knowledge that you’re going to bring to the table and maybe a unique data set that you’re going to start inventing with and so like we’re going to see people start inventing with these new things, creating solutions to problems most folks don’t even know exist yet. And that’s why they’re going to percolate out of the most random places. And that’s why I see people doing this whole like where’s the value from AI and it’s like and in a couple years you’re going to see all these things Right? We’re kind of in the early phase of the internet where like nobody

51:01 trusted e-commerce, right? Like I was not going to put my credit card in the early internet onto a website. But then, you know, like encryption came out and Shopify made things easy and eBay had like, you know, escrow policies. And so like e-commerce started becoming more and more credible on the web. And we’re kind of entering that phase where like maybe you’re not going to trust some of these agents or some of these things that like have your credit cards or your money or do these things, but over time this is just going to keep evolving and getting better and people are going to invent new systems on this. And then all of a sudden you’re going to see a new Amazon break out and

51:32 be like where did they come from? They were just selling books. Now like they sell everything, right? And so like people are building these systems right now and these are the things that you need to be looking at is like how can I connect all the dots again? How do I become the glue between these new systems so that I can like amplify the results of these like, you know, one, two, three become the sum becomes bigger than the, you know, the the the the parts. I don’t know. I miss I’m paraphrasing. I don’t know what it is. >> No, that’s totally true. Yeah, and and the ROI conversation is important. I will say, um, you know, when when people are in the token maxing phase, I think there was a

52:04 little bit more question as to some of that stuff and ROI. And you still have to think about ROI, but like some of the use cases we see just like either save or make so much money, it’s unbelievable. Um, and there’s just so many there’s so many inefficiencies in the world that can be solved with better technology, right? There’s so many there’s so many engineering problems out there just like waiting for somebody with a little bit with an idea and some some grit to go and solve them. There’s just so many of those. >> Yeah. High agency. We talk about the high agency, is it, yeah.

52:35 >> And to your earlier point, there’s a lot of companies where that’s just not the case. And I feel like especially like the the larger, the more the most established, the less likely you’re going to have those folks that are tinkering with AI and LLMs and get a sense of what you can do and then even less likely they’re going to be folks that can actually drive that forward and stick their neck out and say like I’m betting on this new technology, I’m taking a hit on this, and I’m going to move forward. Um, so um, yeah, totally. Like I think it’s wide open. We’re I think everybody here in the chat, everybody here on X

53:07 probably the 1% right now of early adopters. Uh, and like uh, Levels here is probably like the 0.001%. So maybe he’s looking ahead to where you have infinitely customizable software, but the vast majority of the population is nowhere near this and you’re just going to have years of opportunity where you can come in and help these companies, especially the more established ones. And also that’s where the most revenue is to be gained as well. Facilitate this with custom software and getting on this bandwagon. >> Yeah. Well, and I just think like

53:39 I made the like I I had a post here that was a response to this that went a little viral because you know I was like post regularly about how much money makes with custom software launches new apps build software daily with AI tells everyone software is dead but keeps building software right and so this is what I think like sure are people going to build new things are some people going to build different systems and kind of replace some of their stack sure but like not everyone’s going to build these complex apps on the fly and I made the analogy of restaurants right like you could go to a grocery store right

54:09 now buy every possible ingredient you want you know you have a beautiful kitchen at home and cook whatever but like most people don’t do that I just want to go to a restaurant to have like a nice meal you know like I might cook stuff at home I might do that you know but like at a certain level I’m just going to go buy and go to a restaurant and the restaurant might create a really unique experience and I think it’s a good analogy because then you can see like just because a restaurant isn’t like a national chain doesn’t mean there’s not value to be made by selling meals you know in a localized setting or to the network or to the group that are interested in eating your food and so I think like this is the perfect type of analogy yes we’re going to see a lot

54:42 more software software and again some software may be dead some very basic stuff might be dead but you know like I I just don’t know Manny you shared this Andre Karpathy post here too >> Mhm. >> Ethan actually I’d be interested to know do you have thoughts on software being dead beyond this here and just yeah we can we can just kind of >> Yeah. >> and we’ll go we’ll go we got we’ll go back to the field engineering thing here but I think this is interesting because it ties back into like what’s happening here and the opportunity for professionals who are kind of building their skill set here to like just understand that this opportunity has not

55:15 passed you by like we’re just starting to see the beginning of this. >> Yeah and I and I think I will say I think you both have like like both have interesting points here it’s like software is like certain types of software is dead or there’s certain types of software that will not be able to be sold like as SAS companies anymore right um the other day I had a a where I was doing something I hadn’t done in a while. I was doing some like the row machine stuff. I made an app for only that workout. I’m never going to use that app again, you know? Like there’s like that’s how customized software can be for you. You can make

55:46 stuff that’s like one-off for a very specific thing that you’re doing. Just made it in Replit, published it, used it for 25 minutes and then I’m never going to use that ever again, probably. Um but at the same time, if you do know a lot about a problem and you can make something that’s much more deep and advanced, like there’s no reason for somebody to to vibe code, you know, something that like they don’t know anything about and then they that they just like it would just they would just benefit from the expertise on. Um so in that sense, like software is just going to continue exploding. Um and then how it interacts with agents is also

56:19 something that’s going to be more and more important. And already is so important. Uh how your software interacts with agents is is just so so important. >> Yeah. I also make the analogy of like we’re kind of entering the um like if you remember the early blogger days where everyone was just kind of spitting off content and you know, like everyone became like it when WordPress came out and kind of uh putting things on the internet came out, all of a sudden like all the traditional media and all the people who were kind of established publishers were like, “Uh look at all these like bloggers over here. They’re

56:49 not really media. They’re just people tinkering. This not really important. We are the established like publishing, you know, uh megalith.” And then you started seeing people like just publish blogs and build websites and build little interactive experiences and maybe launch a little uh login to that and just started tinkering in that space and that kind of exploded. And like you’re saying with disposable software, sometimes you didn’t really care about like if people read your blog and then sometimes they went viral. And so sometimes you might put a an an hour or a day into writing a blog and sometimes you might spend weeks like doing this like deep journalistic thing that you maybe want to uh hang out

57:22 for a while. And so like again, I think we’re in this just like shooting ideas out into the world and blogging was the first first version of that where I could just take text and throw it out into the world. This phase is kind of like blogging, but you’re writing with code. Instead of writing with like English language, you’re talking with English language, but you’re bringing this idea to life. So, instead of just a narrative, you’re actually building this interactive thing, but it’s the same cycle. It’s the same idea where you’re just empowering more people to put their ideas out there, and these ideas are just going to get interesting as you build, and like you said, you bring in the unique skill set, you bring in the unique data set, you bring in a couple

57:55 interesting connectors, and now boom, you’re off to the races. And so, as you’re a business, as you’re thinking about this we go back to like how many Blockbusters are about to get Netflixed here, right? And how many people are just not going to like see this happening, and this is where again, the opportunity comes up to I’m not saying we’re not going to shift or that jobs aren’t going to change, but the idea that software is dead to me is just so foreign. >> Yeah. >> Um re-routing things a little bit back to our topic. Think I guess I’m I’m wondering um as you’re going into businesses, what are some of the biggest objective objections that you’re seeing?

58:28 Like you have a champion internally, everybody’s raw raw, they’re excited about rolling out Replit. I’m sure you have some folks maybe from security, from IT, or from else elsewhere in the organization. Like what are Talk to us about some of the big objections that we’re seeing and how are we overcoming them? >> Mostly security and governance, right? Companies don’t want their employees to create things that will be shared externally and share, you know, internal data externally. They want to be able to um let their employees have kind of a sandbox, a sandbox environment where

58:59 they aren’t doing things that uh would, you know, not be valuable or be detrimental to a company. Uh so, a lot of the enterprise functionality and features that we have uh helps with that. It helps set up, you know, templates, makes, you know, certain types of publishing restricted, uh can even like limit the types of models that you’re using and things like that. So, we’re we’re working with them on all those different pieces that they need to be successful on Replit, and then making sure that we implement that for their kind of custom use case. Every Every enterprise is very different.

59:30 You know, they’re they’re they’re very they’re working on different things. They have different people leading them. So, we kind of have to customize how they use Replit in a lot of different ways. >> Is there something about Replit that allows you to do that where people can’t get that elsewhere? >> Yeah, we just have a lot of great enterprise functionality. We have a lot of great enterprise features. We have kind of you know, Replit is a company that’s been around for 10 years and it’s kind of like the first I don’t know how however many years 7 years of it were spent just preparing for this moment,

1:00:01 like building the infrastructure for this moment. You know, like after agent was released, that was kind of like a huge deal for Replit as a company. But before that we were basically building the infrastructure for agents to use. Like we didn’t really know it at the time or the people working at Replit. I wasn’t I wasn’t I was using it, but I wasn’t quite working at Replit at that point. But didn’t know it at the time, but they were building got kind of like infrastructure for that. So, a lot of that is just kind of carried over and then we’ve layered additional things on top of that. I don’t I don’t know if you all talk about this, but internally we we have this saying where we

1:00:34 kind of like embrace pain, you know, seek pain. Yeah, seek pain. Um and a lot of these things require huge rewrites, huge architectural changes that really are hard to do and it’s hard to dedicate a ton of time on something like this because you’re not like releasing it in your feature releases. You’re not like oh hey, we just re-architected huge parts of our you know, product. Nothing’s changed for our users, but it’s made it way better and scalable and give us a lot more flexibility of things to build and do in the future. That doesn’t show up in a

1:01:06 change log necessarily, but it’s super helpful for enterprises and that’s kind of what we what we’ve done. >> On that same line, can you help me think through like how you’re approaching requirements gathering, prompting. I know a lot of people, um, like I think they underestimate the amount of thought that goes into what happens before they start sending the prompt, right? Before that first prompt, talking with the project maybe, asking the agent to explore. Uh, I still use, you know,

1:01:37 we just introduced chats where you can kind of go and chat with, you know, your different projects and you have, you know, context across the different workspace projects. Um, so maybe tell me, and even when I’m in the project itself, a lot of times like I have a, uh, LTTT is my short code. Let’s talk this through is what I tell the agent. And it basically like injects, “Hey, before you write any code, let’s talk this through and investigate what’s the best idea and how to look at this, right?” And so like often times I’ll just ask the agent. So like help me think through this thing. So what are you doing with the client? Like when you reach them, Ethan, like you know, like maybe maybe we can kick this

1:02:10 over to you, but like what are you helping them think through? What are you organizing before? How much time are you spending putting the prompt in? Cuz I’m sure you’re not just saying like “Connect this to Google Docs and plug it into my thing.” Like sometimes we do in the demos, right? Again, because we don’t have the strength, [laughter] right? So we’re just like, “Bring in my calendar, bring in Slack, do the thing.” You know? And so like you can’t really do that a lot of times when you’re dealing with these businesses. So how do you think through that? Like how do you think what they need to do, what they need to organize, and then that first

1:02:40 prompt? Like how do you approach that when you’re going to implement this stuff in like a real way? >> Yeah, that’s a great question and, um, I think my favorite are the opportunities where I’m actually talking to them about a business problem. So, um, I I recently had an experience where they were holding an office hours and, um, the customer had brought in a bunch of people from various departments, um, various teams, and basically had a like

1:03:11 a almost like a like a vent session of like their problems. And they had invited me as kind of like the last the last half for the last half hour I was going to showcase Replit. And I just had my note taker on and I was listening to some of these problems and then after one one guy he he had such a great description of exactly how long it was taking him to do a particular task and and he was saying that this was that

1:03:46 that there were thousands of other employees that were doing this particular task. I actually just took that transcript threw it into Replit agent and while the rest of the people kept talking I built the application live like in the background while he was doing that and once it came time for me to present I just showed them, hey this is what I did like I just built the thing that you said would save you hours a week and this would save thousands of employees hours a week. And

1:04:21 it it like literally blew his mind and and I think that to kind of answer your question it’s finding people who really understand the problem and getting them to talk about it, asking them questions, getting them to follow up on like what are the actual actions you’re taking, what are the what is the the painful thing that you’re not able to do or what is this taking you away from that that really helps to to give you that that context that you can then give to agent.

1:04:56 Um So so I’ve I’ve seen a lot of enterprises they’ll have like a team meeting and they’ll talk about a problem and then they’ll take that transcript have Replit agent kind of condense it into kind of a starting prompt and then they’ll they’ll send that off as the first iteration. And it And it works. Yeah, like literally yap mode. >> 100% >> [laughter] >> People underestimate that you could just kind of talk into this thing and then have it clean up your ideas and then create a plan and be like then read the

1:05:28 plan and then be like is this actually what we want to do? Yes, no, let’s modify the plan. Let’s you know, and then go back and forth and like just don’t always take the the AI’s like first output or this first thing or like you know, that these systems are kind of made to just like agree with you and so you know, like yapping, talking through the use cases. When a user comes into my system, I want them to be able to do this. We have this existing system that works like this. We want to bring this new feature in that allows them to do this with this but still do this. How do we best do that? Yeah, right? And now you’ve kind of let the agent pick these pieces, present me the best solutions.

1:06:01 Hey, how do we make sure this is performing? How do we like optimize this? And so like again, thinking through the prompt. Maybe I’ll I’ll I’ll send this over but I Yeah, let’s focus in a little bit on this cuz I think this is where people could use a lot of help. >> Yeah, and I I always say one thing that we see with uh enterprises in particular is there’s kind of like three layers to this. There’s maybe three things that happen before you send that prompt. So the first one at the kind of highest level is at an enterprise level you want to know like what connectors are allowed, what systems you have access to and are giving people access to, and then what

1:06:32 kind of skills or information do does the agent need across all the people at your enterprise. So if you want to connect to you know, your big query, your data bricks, your um you know, Microsoft fabric or any other type or snowflake, any other type of database, uh you need to have that set up at an account level so your admins have set that up, gotten the roles, accesses, everything like that. And then you also have those skills, right? Do Should agent be writing in C#, should agent be writing uh to connect to these particular systems or not? All that stuff is kind of at the very top

1:07:05 level. Then there’s one level below that, which is either personal or maybe even business unit specific. So, I don’t know about you all, but I have a bunch of internal skills that I use to help me with my day-to-day work. These aren’t really applicable to other people. One of them is like like helping me write an email in my voice. I will say I edit pretty much all of AI writing still or a lot of it, but it kind of helps me give like a first draft going. And it’s all very specific to how I write my emails. You know, a lot of skills like that. That’s like a simple use case, but a lot of skills that are kind of personal that are pre-built. Before And this is again,

1:07:37 before the prompt has been sent. So, though you’ve built these up over time. And then the last one is that, you know, you’re going on a I like to say I like to rant and rave to agent. You know, I just like turn on voice to text and I just go and talk exactly about my problem and just go go go go go. So, all these three things are happening. You have your kind of overall architecture, you have your connector set up, your governance on those. And then you have your individual skills that agent can load at any time. And then there’s that context piece, which is where you’re doing that first prompt, making sure you understand exactly what agent would be building, understand exactly what you

1:08:09 want to build. You know, a lot of times you don’t even know exactly what you’re building until like you’re kind of into it. So, you need to go back and forth with agent. You need to talk it out to get to get a full understanding. >> Yeah. And to to add on to that to to kind of go deeper cuz we talked briefly about this Rimar about how like a lot of people first prompt, okay, so maybe they get their first draft and it’s not exactly what they’re looking for or maybe it’s like a a tool that they’ve built up over the last several months and it’s kind of a

1:08:40 larger project. It’s like those principles still apply. So, they’re like kind of just talking it out with Replit agent is still super is I mean, very if not more useful in a project that is already existing. Like one of the things that I will often do is I it go through and give feedback on different pages. So, I’ll just I’ll just go speech-to-text and I’ll just kind of rant about um hey, like I don’t really like this page that much.

1:09:12 It kind of feels too clunky. Like the um the cards are placed in a weird way. I want to see more information at a higher level. Um I want to see a feed in the bottom right. And then I’ll just I’ll just like talk through what I’m experiencing um with Replit Agent. And then I’ll kick that off as as a as a as a prompt for Replit Agent to kind of help me think through what the solution might be. And then

1:09:44 um especially with like Canvas, um I I think that’s a a super powerful feature where you can like literally just create dozens of mock-ups extremely quickly um to kind of decide the direction of where you’re you’re wanting to go with it. >> Yeah. Yeah, and I think that’s um the speed of iteration, the ability to again, I see a couple people talking about the LTTT, Let’s Talk This Through. I think a lot of people underestimate the amount of time and effort that you should spend kind of thinking about what you want to

1:10:17 build. Part of like uh the development process is like uh requirements gathering, right? Like a big piece of the process, if you’ve ever like built software before, before you build anything, before you start writing any line of code, before you start like just anything, right? Like so, in the AI world, we just kind of like go into the agent like just build some stuff, right? And so, you like it gets started going to build, but in like this level in the pro world, part of what you have to do, and we talk about this whether you’re a consultant, whether you’re an agency, part of that initial sales process is

1:10:49 like asking, what is the problem we’re going to solve here? How are we going to approach this? And then that lets you understand how do you like look at their system, identify what you need to plug into, what connectors you might need to bring into bear, and then again, hey, how do we do this? Maybe I have to do some research and find some documentation that I need to feed to the agent, right? Maybe I need to organize some kind of specific skill for how to look at this. Maybe I need to pull some kind of external data set and scrub that in a certain way so that I’m not cluttering the the agent’s context

1:11:19 window with like data that I’m not going to use. And so like there’s all these things that you might need to think about before you actually send that first prop. And that’s where I was, you know, like kind of you know, maybe maybe dive this a little How do you narrow that in? How do you know what’s too much, what’s too little, how to get too com- cuz you can overthink it at the beginning too, right? And then you can like throw too much at the agent. So then I think, you know, breaking down that scope into like digestible chunks is then the next step, right? So like if the first step is understanding what to build, how to break it down is kind of the next step, right? >> Yeah. And And one other thing I will say

1:11:50 is um you don’t you don’t need like your requirements gathering process or your context gathering process doesn’t need to be devoid of any building. Like one thing that I find useful is I will I will have an initial idea, I will go and build a very rough outline or prototype of what I’m thinking things should look like, and then I’ll show it to people, I’ll invite people to the project, I’ll have them kind of take a look at it, and then they’ll give some feedback, we’ll think it through, and then I might even start a new project depending on how

1:12:22 deep we how big the changes we want. But like software is so accessible, like you can use it as your kind of initial context gathering, right? And you can kind of build bring it into your process. You can use agents to help you with gathering that context that you need to give to the agent to finally make the kind of big final production system that you want to push. Yeah, I think this is something developers do all the time, right? Like a lot of times when developers start tinkering, right? So in the earlier days maybe you would have built a prototype in Figma. Or you would

1:12:54 go to some kind of like prototyping tool uh like Proto.io or like you know like some of these you know I’m thinking like no code era and you would build some clickable prototype, you know, like uh InVision. Do we all remember InVision and like you know these like super interactive prototypes that people would build, but they were like totally useless. You would hand this over and now the dev team has to start fully from scratch and they don’t really understand like what happens at different screen sizes cuz this is not responsive and this look You know what it’s like? Now, like you’re saying, we can throw a framework together, we can throw a prototype together, we can go and test this thing really quickly. We can get buy-in from stakeholders. So, you’re

1:13:26 basically saying like not only am I field engineer, but I’m a project product manager at a certain level. I’m the product owner, I’m helping them walk through, I’m the I’m the co-inventor. So, again, you’re bringing so many different skill sets to the table as you help them identify and understand this. So, like That That’s what I think is interesting here is that like you have to kind of see different pieces and then almost act as as a translation layer, right? Understand the business requirements, understand the opportunity, understand the tech technology, bring them all together inside of this like, you know, application that solves again the user’s problem. Going back to how we started is identifying that problem and

1:13:57 I ideally solving it with this tool. >> Yeah, and there’s two things that come to mind from what you’re just saying. So, one is that um now like product managers are so empowered by AI because they already have a lot of that understanding and that kind of like knowledge of what needs to be built. And what they built as a as maybe a prototype that maybe doesn’t go to production. Sometimes it does, I will say we make that like a lot easier when you’re doing it in code and you’re actually like able to fit the constraints. But let’s say it doesn’t. Uh even that prototype can be so much more detailed. At it’s basically not a prototype at that point. Like because

1:14:29 it’s connected to real data, because it’s actually like integrated within the product or like using the same code base as the original product. Um the prototype fidelity is just insane. It at that point it’s like basically a minimum viable product. You know, it’s like you you could almost just push it. They just want to do a couple of changes. The other thing that we’re seeing is that um historically, you know, you you would have your product manager, like they would interview people, they would gather requirements, and then they’d work with their they’d tell their software engineers, “Hey, we need to to build this.” Uh and the software engineers they would like kind

1:15:01 of consult with the product managers to understand those requirements. They’d have to go back and forth, etc., etc. One thing we’re seeing a lot more of is software engineers or those with kind of the technical expertise are now acting in the more consultative role. So, um a a product manager might be able to build something in uh a tool like Replit and then actually make it good enough to and it’s, you know, would have all their governance controls set into it. They’ve got their right templates and everything like that. And and get it all the way to the point where it’s ready for, you know, production. Uh but the software

1:15:35 engineer might come in as more of like a consult like, you know, again, they’re like consulting with them. They’re being like, “Hey, one thing you should think about in terms of the, you know, the way that we have our tech set up is you want to do this or that.” So, now they’re kind of like sitting Software engineers are sitting at like a higher level now instead of uh historically where they were the ones just taking the requirements, they’re like consulting with those other people. >> Yeah. >> And and and one other piece in that is like the designers as well, right? Like, they would often create these these prototype mock-ups. Um and it’s it’s almost like you’re collapsing these these roles a bit into

1:16:10 um kind of more of a builder profile where it’s like uh a designer is now able to to make some amazing prototypes um that they can hand and work with the product manager in. And then the software developer can also work in the in the same on the same thing. So, it’s like um one of the the really cool things that um that Replit has unlocked is is actual parallel development. Uh which uh I think is so cool to see when um when

1:16:42 teams are working together on the actual same product. Um >> Say Say Say more about that. >> I was going to say, tell me more. >> Say more about that. It’s a big deal. And like who exactly is working together? Is it engineers working together? Is it technical and not technical people talk talk to us more about that? >> Yeah, so just like Axel said, like there’s there’s kind of an interesting bringing together the product managers or the person with the business problems and the software engineers or more technical folks who kind of understand

1:17:15 how systems fit together. And and so what we’ve seen is you’ll have somebody who has a business problem. Um I’m trying to think of a way to abstract out some things. Um Like there is a great use case where somebody was using a particular SaaS solution. And they had difficulty with using its API because it didn’t have the ability to first of all have great visibility

1:17:47 into who’s using it. Couldn’t rate limit them very well. So that they they would like use up all their quota in a day for the week. And they’re like, “This isn’t cool because I don’t know who’s using it. I don’t know like how to to fix this.” So they they they started using Replit as essentially a wrapper around this this SaaS solution to get access to their data better. And

1:18:19 they brought in a technical guy who was more familiar with like their SSO and all of those things, integrated that into the application so that all of their users that were going to be using this application were securely logging in. And then the business the the kind of the person who um had this problem was able to kind of create the user experience such that people were able to very easily um access this application. And

1:18:52 and it like it was so fun. I was cuz I was this was um I was kind of involved with the team that was building this. And um doing a little bit of consulting, but they basically handled the it themselves. And I saw their their communication thread um where they were going back and forth like the it was this was in a Slack channel. And they were like getting so hyped about the fact that they were collaborating together and building together. And one of the one of the guys was like, I should be able to like do emojis to this guy’s prod. This guy’s

1:19:25 killing it. And like like integrating the SSL, he’s like, oh, this is so sick. And and it’s just so funny how we’re getting to this place where um it it feels very collaborative. Like it used to like we didn’t I don’t think we truly understand how isolated these different pieces were, but Replit is bringing everybody into the same place. Um and maybe there are some like some things like that the the developers do that’s a little bit different than

1:19:56 um than normal folks. But like we’re we’re collapsing into like, hey, let’s like let’s all just be working together instead of kind of being more isolated as we were historically. >> This is a super under appreciated function of Replit is bridging this gap between non-technical people and technical people. So if you’re a non-technical person, you’ve probably always felt limited. Like I have this idea, I want to bring it to life, I want to be able to contribute, but you’ve been intimidated by talking to the engineers. That’s no longer the case. So Ethan and Ash, I’m wondering like so we

1:20:27 know what the non-technical people get with this. They’re super empowered. They have an idea they can bring to fruition. Maybe talk to us from the other side. Like, what do you think technical people and engineering people, how do they benefit from this bridge and like working together in this collaborative way? >> Yeah, we get a lot of feedback from software engineers at companies who have implemented Replit that they absolutely love that they can now just receive code instead of screenshots, that they can interact with the Replit environment in a super smooth and easy way, right? Like, we let again, Replit started out as an IDE. We have a lot of

1:21:00 infrastructure that lets you do things like SSH or communicate directly with the Replit environment from your own, you know, local IDE or whatever you whatever tools you want to do. We give you access to the shell and we give you access to a whole bunch of things like yeah, not to get too in the weeds, but we let you like run Docker directly in Replit. So, a bunch of engineers love that this environment is so flexible. And then, they also get to have they’re not like the bottleneck anymore. Their their plate isn’t just full of like little things. You know, people that the

1:21:32 the business users, the experts can handle the little things. What their job is now is being consultative and then like using that code that they can kind of um bring to production. I think Ethan, you you maybe have this story about the the guy who the software engineer who asked for the Replit. Can you maybe tell that one? Yeah. >> Yeah, I I I wasn’t directly involved, but basically there is somebody who was talking to um an engineer on like they they basically wanted to

1:22:04 incorporate something and um and they were just they were just talking and talking and he’s just like, “Dude, just stop. Like, just send me a Replit. Like, just send me a Replit and then we can go from there.” Like, and and I think that there’s something so cool about that where it like you were saying, it’s just like you’re able to um you’re able to be working off of the same the the same, I don’t know, document or whatever. >> You’re speaking the same language,

1:22:36 right? >> Exactly. Speaking the same language. >> Software engineers are used to working with code, and now everybody’s building code, right? So, if you can if you can take what you want and um give it to people in the exact code that you want it implemented in the product, oh my gosh, your life as a software engineer gets so much easier. >> And even if it’s not exactly that translation layer, like as a guy who spent so much time trying to build software and get ideas out of my head and into a developer’s head and have them try to build what I see, it’s just

1:23:06 almost impossible, right? Like it’s impossible to spend the amount of time looking at the nuance, like thinking about every little detail that you want, right? Because especially when you’re paying somebody by the hour on some kind of fee, they’re just looking at this kind of scope sheet and they’re just like running through it. Especially if you’ve got some like, you know, budget team or you’re doing some overseas thing and like you’ve translated your product requirements to somebody who is maybe really smart, but then they took that conversation and they translated it to somebody else who’s now going to distribute it to the team who’s going to build, and now you’re two or three layers away from like what the problem is and what your idea is. And then you

1:23:38 have to wait two weeks for them to bring you back something that’s like so far away from what you actually wanted, and then you have to okay, no, but we really wanted this and move this over here. And now to be able to just go and do that and get that first thing, even like you said, even if it’s not production-ready, even if it’s not like, “Hey, we’re going to bring this into our enterprise system,” at least those devs understand like, “Hey, when I click here, I want this thing to open. When I do this, I want this, you know, so the logic, the bones of it, the skeleton is there in a way that they understand, and you’re no longer just trying to like describe these abstract concepts. They have

1:24:08 something to work on that’s like fairly tangible. >> And it’s a good filter mechanism for engineering. So, to your point, like do just send me stop yapping about it, just yap it into Repl.it, send me a prototype, and send it my way. So, like if you’re engineering, you’re inundated by all these ideas, like you’re the idea guy, go flesh it out, take the first pass, and then send it my way, and then you can see like which ones of those you should actually take action on. >> And and to to rhyme with what you were talking about, like it takes 2 weeks in large enterprises, it can take 6 months to get a first rough draft. And we actually see this

1:24:43 with a lot of larger enterprises that are adopting Replit. It’s like the the the loss in communication um that happened previously just it it basically negatively impacted everyone because they just nobody was working on the same the same thing. Like they they didn’t have like a north star. And with Replit now, you’re able to build a prototype that’s like you can actually click around, do things. You can you can do user testing

1:25:13 it’s basically in a week, like pretty easily. And that means that you can immediately start getting feedback um and then use that to inform kind of the larger product moves that you want, or uh maybe you want to incorporate a new product line or something like that. You can just actually use Replit as like that um that just like s- kind of the accelerant that makes that process, that feedback loop so much

1:25:46 tighter um that like for large enterprises that it’s just like it totally transforms how they’re able to to to operate. >> Yeah. And I will say we’ve been talking a little bit about prototypes. More and more enterprises, and we have a huge focus on this like are taking their applications that are built by those non-technical users and just putting them directly into their production applications, right? Because you have so much flexibility with the platform before because you can go very deep, you might have that consultative role of the

1:26:18 software engineer to double-check things, how it’s integrated with systems, is it conforming to all your requirements. But man, you can go far. Like we have you know we have small businesses and startups that are running completely on Replit, like using it completely as their only platform of choice for technology. And I think we’re going to see that it’s more and more common, right? Even in instead of even at some large enterprises, like instead of just making prototypes and sending those over the the fence to to engineering, like we’re going to see people taking it through the entire process, getting that that thing really

1:26:48 smoothly built. And this is this is you know, it’s going to change a lot in the in the next couple of months, I think. It’s just going to continue to change a lot. >> Ethan and Axel, can I ask like so we have folks in the audience that are going to become field engineers like in their own respect or some version of that? What lessons do you know now that you maybe wanted to have known as you started this type of function, this type of practice? Like any pitfalls, anything like you’ve run into where like damn like probably shouldn’t have done that or any deals that went bad where you’re

1:27:20 like man, like I wish I hadn’t done that. Like any like tidbits you can like arm people with as they’re trying to do this for themselves? >> Ooh, that’s a good question. >> Like too many. >> Very [laughter] good. >> Yeah, like there’s >> Yeah. >> PTSD like oh don’t remind me that. >> A billion things. Um like I guess one we’ll get we’ll get more specific as we continue thinking things, but one broad one is that you you need to always be like building and tinkering and deeply understand exactly like the product that you’re working with and then also the business that you’re working with. Um one thing I like to like if I if I had a

1:27:53 wish, if I if I a genie let me have a wish, like one thing I’d love to be able to do is like hold an entire company and all the actions that its employees are doing and all the processes they have in my head. To just be like able to see an entire business and how it’s run. And you know that’s it’s so complex like how it like there’s so many systems interacting with each other, people interacting with each other, processes interacting with each other. But if you can like marry the technology and those processes and businesses uh business processes and like requirements and things like that, it’s it’s amazing. So, do not be afraid to go deep on a particular business. Deep. And

1:28:28 I think this is an undervalued thing. We talk about this. This is exactly the premise of this like demo dilemma that we explained last week where, you know, sometimes you have to get into the weeds and iterate on something. Sometimes you just have to build like the thing, get it into play, and then like does it work? Does it not? And then I’m going to fiddle with it and I’m going to see do I make a little bit better improvement today? Does it get me a little closer to where I want? And it might take you a few weeks, a month to really like use this thing and actually figure out what you need to build. And I think this is like a really good takeaway here. Manny, you asked about takeaways. It’s like

1:28:59 lots of folks I think want to build big complex systems right up front without understanding how they work because you can just keep building features. Rather than identifying, we talk about first principles thinking, what is the first thing I can solve here? What’s the first like thing I can start building and tinkering with that then I can start slowly building into this thing that I want to have like evolve. And I think a lot of times that’s how good software evolves. That’s how good anything evolves, right? It’s like you you you have the core of the idea, you get something into to the real world, you get feedback, and you start slowly iterating and making it better. That’s

1:29:32 the agile development process, right? Like that’s kind of the standard development process in in in up until this age agentic era. So, I think like that iterative process that starts small and expand, add complexity as you go rather than trying to start with all the complexity is a big takeaway here in my mind. You know, like that’s anything I build that I’m still using that I started let’s say months ago or, you know, uh even a year ago. I have projects that I still like expand on. So, those are some of like my most robust things because I spent all this time like just

1:30:03 you know, and that’s what we’re trying to expose here a little bit is that like it’s different to think about that than like from start. So, Ethan, I see you nodding your head. Uh we’ll we’ll give you guys each a little bit of just a round robin here for some closing thoughts. Um take it wherever you want. Field engineering, uh the stuff we were just talking about. Give you guys each just go around the horn here and you know, just just some thoughts about the whole space and and the conversation today. >> Yeah, I think um I think it’s just like maintaining your curiosity, I think is so important. It’s it allows you to keep energetic in such a

1:30:34 like the the the pace of change is so insane right now. Um just like every week there’s a new model. Axel and I were talking about this the other day. It’s like we’re so we’re so busy now at work. Sometimes it’s hard to even for us keep on top of everything going on in the industry. And and I think that um the only way to really to really do that is to have to try and have fun while doing it. Like to make it so it’s not an energy suck. To

1:31:04 make it so it’s not like a like things happening, whatever. Um it’s to to kind of keep that that that uh kind of curiosity, that that enthusiasm to to just keep learning. Um keep developing uh your own skill set and and understanding how to work with the tools. Um and I think that that you’ll be really positioned well to to to be successful with all of this going on. >> Yeah.

1:31:36 And then building off that, you know, you want to work with the agents that you like, right? But you also want to work with the people that you like. So, I’m I’m very fortunate to work with Ethan uh at this company and you all too as well. Uh it’s really helpful to to have that fun like when you are building like you know, we we work a we work a lot of hours, you know, like >> you talking about? [laughter] >> We spend a a of time doing work things. Um and so it’s it’s it’s really hard to do that unless you are working with people that are awesome. Uh and then

1:32:07 also working with technology that’s awesome. So like having that curiosity, having that love of what you’re building and what you’re working with, and then also working with great people is uh it’s a great way to do things. >> Yeah. Uh undervalued advice there, I think. Work with people you love and yeah, just get in >> [laughter] >> get into the get into the weeds a little bit, right? Like I think those are we talk about like um something you’re passionate about, something you’re interested in. We You know, software dead, recreating a to-do app is for sure dead. We don’t need any to do any more

1:32:39 to-do apps. Y’all just stop with that What we might need is like again, your unique take, your unique spin, even if it’s just a unique experience. We talk about like um okay, in my mind community is the moat. We also see in addition to this like software’s dead, like what’s the moat, right? Because if software, if the technology is just kind of prolific, then what becomes the moat? The network is the moat. Software is just the tool for harnessing the network, right? And so like your skill set, your thing, your ability to generate an audience around this thing that you can sell becomes the moat. And you can scale that as much as you want. We showed you the other day somebody

1:33:11 who’s doing hundreds of thousands of dollars managing ads inside of Replit. Uh we interact with all these different enterprise clients and we see these logos that sometimes we can’t even tell you about that are doing things inside of Replit and you’re just like, “Oh my god, what?” Right? And so like I just think that we’re again, to beat a dead horse, at the beginning of this and your ability to learn how to translate these skills, these learnings, these things into practical ways in your world, forget about scale, right? Having a milli like building software used to only be financially viable if you had millions and millions of people because of the scale that was needed to build

1:33:42 the software because it was so complicated and expensive to build the software. Now you don’t need millions of people to make your software successful. You could build software for 100 people and and have a great business. You could build software for 1,000 people and like retire next week. Like right like you don’t need millions and millions of people to make this thing like you know or to build a viable business and I think that’s what people need to understand that the shift is coming from here. Like you know, like this I’m not saying people won’t break out and build these mega systems, but I think the next layer of where lots of money’s going to be made and where we’re going to see tons of opportunities is basically this type of thing, people building these

1:34:15 things and sprouting out with you know, these small little nodes that then become big nodes. And so anyway, that’s you know, some some some thoughts on my end as it relates to that and what’s coming. >> I think it is really cool that all of us in some way we’re tinkering with Replit before we ended up here at Replit. Isn’t [laughter] that funny? So going back to your point, I think actually you made this point or Ethan around like following your curiosity and tinkering with stuff and I think Ethan you said like find the agentic platform that you like and work on that. And then Raimar you said like build with others, find a community and I think we’re all examples

1:34:47 of that, right? We started tinkering, we found Replit, we liked Replit, we started building with other folks and you just don’t know where Like now here we are like here on the inside helping to guide the product and change the the the product from within. So I think for me that’s like the overarching advice here. Like in this agentic era, find the tools that you like, find the community for those tools and just build build build, build in public, build within your company. >> Yeah, and build value amongst the network itself. Don’t just build the software to build the software. Right,

1:35:20 the software has like software may not be dead, but software itself is not bringing the value. Right? Like so in that context, this idea that you’re just going to get the value by building the day of like if you build it, they will come is just over. Right? So like now you have to build something that either applies practically to a system that already exist, a business or whatever, or use it to create some kind of unique opportunity that you can leverage to then grow or scale and bring value to the software. Um so again, I think Replit becomes a really powerful tool to do that from idea to production at whatever scale you’re at, right? Whether you’re just getting started or

1:35:51 whether you want to work with a large enterprise. So, anyway, uh Axel, Ethan, thanks so much for joining us today. Any final thoughts as people go? Where can people connect with you? Uh I think we have some of your social profiles up here. I think these are your LinkedIn handles if anybody’s looking. Um I think uh you guys are a little more active on LinkedIn than you are on X. Uh but >> I’m probably actually more active on X, but I’m not super active on social media generally. >> Got to say. >> [laughter] >> But maybe I should be, you know? >> Got to get in the game, man. >> This is what we talk about, yeah. >> Yeah, for sure. >> love to connect. So, if you want to connect with either of these folks, feel

1:36:23 free to reach out on X or on LinkedIn. I’m sure they’d be happy to connect. Make sure you send a note. Don’t just connect. And then please don’t just try to sell them stuff or like ask for help or like be generally like, you know, just internet person. Like just be, you know, “Hey, saw you on the stream. Like what you had to say, you know, maybe we can connect.” That’s it, you know? >> Here’s a good tactic. You should take the biggest thing that you learned from Ethan and Axel here and just post about that, share that, and tag them. I I love those It’s easy, you know, like here’s Here’s my Here’s my biggest thing I took

1:36:54 away from Axel is a great cuz here’s my own take on it. Here’s where I’ve seen it played out. Thank you very much. That would be a nice thing to say. >> Love it. >> There you go. >> Man, it was it was so great being here. This studio is so awesome. It’s great hanging out with you guys. Like this is great. >> Yeah, just um just coming together real nice. Like it’s starting to improve. If you’re watching out there and you’re seeing us kind of just tinker, right? You’re watching the same thing. We’re like uh you can’t really test this stuff without an audience without going through the steps. So, here we are testing it. You see week-to-week we try to improve it just a little bit. And over time you’re going to look and see like, “Wow, where did these guys come

1:37:26 from?” The same way when you’re building software, right? It’s the same process. Slow iteration. Uh what do they say? Like uh 10-year overnight success. You know? >> Yep. >> A lot of times the work that really gets you across the edge is not things people see until you break through the surface. So, anyway, if you’re out there, you know, work hard, keep pushing. Um maybe you haven’t broke out yet, but your moment is coming. So, we’re going to let Ethan and Axel get out of here. They got to go back to work. Uh Uh and yeah, Manny, you and I can stay here and take a look at the comments and just see what’s kind of coming in. But, thanks, guys. We appreciate you so much. Uh we’ll do this again real soon.

1:38:00 >> Yeah, I mean, I just love like every so often on a regular basis I get like a introduction for a new Replit employee. And my favorites are when like they were members of the stream before coming to Replit. So, like, “Hey, Manny, like I actually in the stream chatting and talking and viewing or now like I’m here at Replit.” I’m like, “Oh my gosh, that’s crazy. Like, that’s so amazing.” And I think I think you’re an example of this, Raymar. I think, you know, the brothers are example of this where like you like you build, you follow your curiosity, you you show it, you showcase it. You just don’t know

1:38:32 where that’s going to lead you. >> Yeah, and they’re getting out of here too quick. Thanks, Ethan and Axel. Yeah, super smart. I we honestly could have probably just talked with those folks for hours. Um I think this is the type of conversations that I think um are really valuable. I know on Tuesdays we like to dive in and watch the agent build and we will do that. Like, we’re not getting away from just building with the agent. Um but I think also having these types of conversations are super valuable for folks and diving deep on the thought process. Again, how to practically apply this in your world.

1:39:02 I really like the idea of doing a little bit more of this. Um one of the things you’re going to see as we get through October. So, right now in October, we’ve got this pro series already scheduled, but I’m already starting to look beyond that. So, in November, we have a couple things already lined up where we’re going to like again build. Uh we did the interview with the Restream folks. We’re actually going to bring their head of product in and talk a little bit about building some of these systems on the back end. And then I have locked in a VC y’all that uses Replit up and down the board, inventor of some very interesting

1:39:35 AI technologies. Uh after we get through this pro series like in October, we’re where just ramping up. So, I think um and and again, Manny, I don’t know uh what you’re planning on Fridays, but I just see this momentum growing. I see us kind of building this little loop. And I think what you guys might start to see a little bit is a little less distinction between like the Friday and the Tuesday where we’re just like, what’s the best content we can deliver? Who’s the smartest person we can bring on to talk to you? Where are we going to point this and do this little segment? Sometimes that’ll be a build along, sometimes

1:40:06 that’ll be an enterprise project, sometimes that’ll be pro series, maybe we’ll try to bring in you know, like some of the leadership in and I I I don’t know. You’re just going to see all of this start to really evolve as we go. We’re investing a lot in media and production in general. So, Manny, what do you what are you thinking as we just kind of make this transition as you see this whole market evolving a little bit and kind of getting to do these these conversations as we go with folks? >> I think I’m wondering when do we go three, four days when we accommodate all the great stories that we have because there’s so much to talk

1:40:37 about. I mean, like we want to do more deep dives, we want hands-on build. We have great people within Replit that can come on and give us the you know, the real nitty-gritty in terms of how we’re building and what’s going on in the space. We’ve got amazing user stories to bring on the show. We’ve got new features to launch and educate you. So, you know, you know, me and Ramit were talking about this where like we want to like just make this like this two days of the week like the best possible experience and really refine it but I think pretty soon like we’re going to go past that. Speaking of which, on Friday we have a builder story. We have Crust

1:41:12 Data that I that I’m pretty sure is going to come on the show. This is a startup that has a new MCP in Replit and they’re also going to talk to us about that process, how to get into the startup program, and then also how they use one of the skills to land a 200K deal. So, going to hear from them how they’re building and then we’re also going to have the folks from our recent Atat acquisition. So, if you notice we’ve been talking about being able to do inline charts within chat, great visuals we’ve talking we’ve seen that on past shows. We’re going to

1:41:43 talk to the team that helped bring that into Replit. That was via Ada, so we’re going to talk to them about like what is the world of data analytics look like in the agentic space and how to do that now as well. So, that’s going to be a really good show on Friday. So, be sure to tune in for that. But, no I think I mean I like I mean I think for you know we’re we’ve been talking about it’s it’s like Replit TV, right? Like we have a segment here, we have these folks come in, we do a live build here, have a new feature to talk about, a new product, new guest, and then just like roll out these segments and just like make sure like these two days of the week are like top tier, like great time, great value for

1:42:18 everybody tuning in. >> Yeah. Well, I just just watch. Um, you know like again, the trying to trying to increase the production value a little bit, trying to tighten up, you know, how we prep for the the shows, the making sure that the guests, you know, are also like ready to be on air. And so, like this is all kind of a new thing. Like most companies would have like a whole dedicated team to kind of this and we’re doing this as part of the community team with like a whole lot of other responsibilities. And so, like we love that we get a chance to do this with you all and we really think this like makes the experience authentic of trying to like build things and then

1:42:50 work with you all and like really closely think about, you know, how we expand and grow on some of this community stuff. We did a town hall just recently. We do a lot of stuff in office hours. Tomorrow we’re going to keep on some of the sales stuff. I’m actually trying to bring in a couple of the sales folks from Replit to see if they’ll come and hang with us in office hours tomorrow to talk a little bit about that as well. So, yeah Manny I I I agree. I think this is going to evolve quite a bit over the next 6 months especially 3 months as we just kind of really you know, we’re we’re we’re building processes and systems, right? Like you

1:43:20 guys see this behind us. We’re we’re just you know, tinkering, iterating. Let’s see. You guys are the this week in tech show for Replit. Yeah, this week in this week in Replit, right? >> [laughter] >> Um, yeah like I think there’s like the systems that we have to put in place like Remar’s doing a lot like in the studio like setting up all of the lights and the cameras. We have a new studio that we’re building out. And then we’re also trying to hammer down our processes for like, okay, you know, who should we feature? How should we prep them? How do we like prepare them to get the best possible show for you? And also make,

1:43:53 you know, make them look good as well cuz like live streaming is new to a lot of folks that are coming onto the show. And so I always love seeing like they’re so nervous about a live stream. Well, how do I prepare? What do I do? Then they go through the live experience and like, oh, like everybody here in the chat is amazing. But also after they come on our live stream this is something we’ve been talking about Remar like we got to like to do better here. Like somebody comes on the show, we should be able to package up their segment and put it like on socials right away. Clips, assets, like just like, you know, get as much value that we had with their time and like amplify that through

1:44:24 media and other things that we want to do. And also like, you know, how do we like build that into our community hub and all this stuff? >> Yeah. >> So, there’s like there’s so much to do. >> to do. >> Yeah. And we’ve been you know we were like I feel like, you know, Remar and I have been talking kind of inside look here is like I feel like we’ve been going really fast, really strong and we’ve built out a lot of cool stuff and I feel like and we both feel like it’s time to refactor things and like bring everything and tighten things up and automate things and streamline things so that we can continue bringing you even higher quality shows and content.

1:44:57 >> Yep. >> Yeah, refactor. >> Yeah, one of the refactor. >> That’s right. >> [laughter] >> Um yeah, I the live thing again, I don’t think we understand how valuable or important this is going to become uh because it is like an authentic touch point with the community. It’s real. It’s not edited, you know, like we screw up sometimes. The demos don’t always go perfect, but then we do get back to building. As Ru says, go back to building, Remar. There she goes. Exactly, you know? Um and so that’s the The thing is like uh we we’ve got ideas of how this is always alive, right? Like

1:45:29 how do we bring this media product to life where like it’s just always alive, just like the AI world is, just like all this stuff. And so like, um you know, there’s limited resources, limited time, but like Manny said, like how do we increase the production value, get better at packaging this stuff up, and then cutting out the little nuggets that are like the the real value, um you know, and packaging that up in a way that like maybe, you know, Ethan wants to share on his social and Axel wants to share, and some of it’ll go on the Replit streams. And so like, again, you’ll watch us start kind of churning this stuff out, and if you want to like

1:45:59 get involved in that, maybe you start clipping some stuff up, right? Like there’s a there’s a big opportunity to take some of these learnings and kind of, you know, break the little nuggets, especially if you’re good at like selecting what is the value out of this like hour and a half, 2-hour live stream. And if you’re good at being like, “Oh, that like sentence right there, or this 30 seconds, or hey, this 15 plus this 15 plus this 10 tells a real cool story here.” And then you post that, you know, there’s a big opportunity there for, you know, you guys to start just putting clips out, and you want to earn some credibility with us or earn some like juice

1:46:31 uh with Replit across the board, start doing that type of thing. Um so yeah, there’s just a lot lots lots lots for us to do here. Um and we appreciate everyone who hangs with us all the way through these shows. You know, the watch time’s kind of crazy, so yeah, thanks for everyone who’s here with us uh towards the end of these things. >> Can I say that, um you know, like for me I’ve always been video first, and that’s just been my medium. Like it for you might be different, maybe blog posting or something else, um or shorts, but my my my my medium has been like this 5-10 minute like uh exploratory video format. But I also found that like a really short quick win was just grabbing

1:47:04 segments from some of the things that I was already consuming and clipping that and just sharing that. So, you know, you’re already seeing here, you already like join us for the live stream, like it’s not that difficult to grab like the video, clip the part that you like, just like cop like drop that in and like, “I I like this part.” And like tag me and Ryoma or whoever was on the show like it’s like it’s actually like pretty straightforward to do and I like yeah that’s been like a really like nice way to also like fill out my content calendar when I haven’t been producing as much as I as I as would have I would have liked. >> Yep.

1:47:35 Yep. Um okay, I think that’s uh any other any other thoughts? Anything to to think about here? I know um that’s kind of a lot to take in so >> Well, I think for me uh one quote that I liked of yours Ryoma or was like a lot of folks are What was it? Like a lot of folks are going to get Blockbustered or Netflixed? What was that again? You said it like twice. I’m like yes. >> of Blockbusters about to get Netflixed. >> Yes, so explain that to us. What does that mean? >> Yeah, so what I mean by that is that um Blockbuster had the distribution in

1:48:08 place, right? They had the place where everybody used to go to get videos. And now like Blockbuster doesn’t exist. I think there’s like one in the Pacific Northwest somewhere, you know? Uh now we all go to Amazon or uh Netflix or you know, whatever streaming platform of choice. You enter your streaming platform here. And so what we saw in the transition was that even though Blockbuster had the infrastructure and they were the place that people went to go like rent videos and you know, like kind of the family weekend experience whatever. Like I still remember going into the store and you would go and like you know, find what video you wanted to watch. It was

1:48:40 like a whole experience. And almost nobody imagined that Netflix or some digital thing was going to come and replace that, right? And we saw the advent of like Redbox and Netflix as a you know, like sending you the DVD, right? Like oh my god, they’re shipping me a movie to watch and then I send it back. But like they innovated and they created a new market and the next thing you know, slowly there’s no Blockbuster. Everyone’s on Netflix, right? And so we’re kind of in the era where there’s lots of businesses that you look at and this is the stuff where people like SaaS is dead. You see these businesses that are just not evolving. They see the era around them,

1:49:12 but they’re not really sure how do I become the next version of this? They’re so stuck. You saw Kodak do the same thing where they just couldn’t keep up and then digital cameras came out like, “You’re being to tell me the best like people making cameras didn’t like transition into this digital era?” Well, it’s because of what Axel said. They can’t. They have systems. They’re already bought in. They’re already the Titanic moving in this direction and the icebergs are everywhere. But you have a small little boat. Your business is nimble. You are a forward thinker and so you’re going to be able to like take these things and move and figure out how to become the Netflix competing against

1:49:45 the Blockbusters. And you will be an underdog and people will think like that’s never going to work for a while, but if you find the right product and if you do the right things and if you like again figure out your distribution, you can overcome. You can be the winner of these battles. And so that’s it. When you talk about blockbustering, you know, getting Netflix, that’s that’s the thing I mean. There’s going to be so much of that in the next handful of years here. >> Yeah, that makes sense. I think I take two things from that. One is if you’re within a if you’re within a larger organization, just appreciate that

1:50:16 innovation is going to be a little bit tougher whatever reason. You might have all the cards. You might have distribution. You might even have the tech. The case for this is Xerox. Um, you know, I don’t know I don’t know if if the folks know, but Xerox had the initial Mac like personal PC idea in-house in their lab. Like they had the mouse. They had the screen display. They even had printers. And um they had Steve Jobs do a tour of this. This is like something that Steve Jobs was able to negotiate with Xerox or something. I want a tour of this lab. I think I he started taking notes, writing down ideas. And it wasn’t Xerox that came out with that first PC or personal computer.

1:50:51 It was Apple. And so even though Xerox had everything, they had the tech, they had distribution, everything, their business was just set up not to encourage that. And as a result, um Xerox got the Apple treatment like Blockbuster got the Netflix treatment. Fun fact here, Reymar. Um in 2000, Reed Hastings and Marc Randolph, the founders of Netflix, offered to sell Netflix to Blockbuster for $50 million. They said no. It was too expensive. This whole thing was not going to work. Today, Netflix is

1:51:23 worth $290 billion and Blockbuster is no longer around. Um so, that’s lesson number one is like if you’re within a larger just appreciate it and I have been in a larger organization as well. I can tell you that’s the case and you just have to like just strap up and like know it’s going to be a tough road long. On the other side, if you’re like the folks that’s looking to disrupt that business, just appreciate that maybe the odds are a little bit more stacked in your favor than you might otherwise think. It’s a David Goliath versus thing, right? Like Goliath looks all tough and big, whatever, but you got the slingshot. You got the tech. You got the edge. You got

1:51:55 like the entrepreneurial like grind and like the ability to use these tools to actually outmaneuver these bigger organizations. So, depending on where you are, I think there’s there’s lessons to be taken from this. >> 100% lots of lessons, you know. Um Yeah, and I think again that that’s not going to play out the exact same way, but it’s going to rhyme, right? There’s going to be lots of similarities. I see so many similarities to like the early web development days where like lots of people started figuring out how to build websites and kind of get online. Again, I go back to that blogger analogy where people just started putting things on

1:52:28 the internet. Some of that went viral. Some of those folks now became the new infrastructure that is journalism, media, new media, whatever. They became influencers. They started YouTube channels. Their podcasts became global sensations or whatever. Niche sensations, right? Again, like just they found their audience and now they’re a whole lot more prolific today than anybody thought they would be 10 years ago, 12 years ago, etc. I think the timeline for this next phase is going to happen a little bit faster, but it’s the same type of thing where lots of folks are experimenting, tinkering, building things that, you know, like might turn into the next Netflix, might not. They

1:53:00 just might be something useful for your family or your friends or like a small group or your business or whatever it is, right? And so, I think, you know, just going back to we’re so early on this. There’s so much opportunity. There’s going to be so much to shake out as this thing evolves. So, um yeah. Uh I don’t know. That’s that’s that’s that’s all I got to say, you know. >> Spot on, folks. Um Yeah, any any questions, comments here that we want to tackle? Office hours tomorrow, Friday showcase. Come on through. Do we have a an agenda for tomorrow for office hours, right? We’re

1:53:33 doing >> Uh we’re just going to keep going with some of the sales stuff and some of this pro level dive in. Um I don’t know. Yeah, exactly. I’m trying to lock in. I got a verbal commit last week from one of the sales guys to come hang, but I got to like confirm and make sure that they’re still going to show up. So, uh we’ll talk a little about that. Um we might build some stuff tomorrow in office hours, you know. Um there’s some things I’m working on from the community hub uh that might be interesting to talk about, you know, with the check-ins. So, some of you all have been checking in. We’ve been talking about what are we going to do with those things. So, uh yeah, just come out and and hang. Uh again, the Wednesdays where we get

1:54:05 really hang hands-on with the office hours and um you know, review different projects. And we probably need a show and tell coming up soon, I think, in office hours. >> Show and tell, feedback on existing projects. Um something to play around with. Like I feel like there we need an opportunity for people to come up and like pitch their product. Like almost like a mini Shark Tank. Where like maybe it’s an office hours where like you sign up and then like you get like 5 minutes to like pitch what you’re building on and people can give you feedback. We can give you feedback and um kind of a good way to test your idea.

1:54:36 >> Yeah. Um this question here I’ll answer just cuz we’re going to bring somebody in to talk about this. But how do you guys think this will impact funding? Sure, might get faster market companies as we’re seeing, but uh another piece maybe people are way more conservative with equity. Yeah, for sure. We’re going to talk about this. I think finding distribution, you’re in an era where like bootstrapping is going to become a lot more prevalent. It’s going to be a lot more difficult to kind of raise money, especially for these like smaller ideas. Um and so again, finding distribution, finding your moat,

1:55:07 uh that type of thing is going to be really really important and um and for the right case, how do you have package it up and actually get the VC dollars, right? How do you position your product? How do you build momentum so that like, you know, when you go looking for VC money, uh you kind of want to have momentum. You know, almost nobody’s going to give you VC money if you just five-coded a little prototype. So, there’s got to be some kind of momentum, some inside baseball. So, we’ll talk about that in future streams, but stay tuned on that. We don’t have any anything defined yet, but I think it’s November 17th, whatever the third I think it’s the third weekend of November, second week of November, something like that.

1:55:38 We’re going to talk about this. >> Mhm. Yeah, it’s so tricky yeah, cuz like AI is so hot right now. So, you probably you have that tailwind. But then like now like software like that you know, I’ve been hearing comments where like software is no longer a moat. Like your software IP no longer exists. Okay, so then what is your moat? And the counters that I’ve been seeing is like your distribution is the moat. Your ability to get attention is the moat. So, it’s almost like you know, you no longer need to have like a stellar tech platform and

1:56:09 IP, but you also have to be moving really fast. So, one advantage I think that we have at Replit is that we’re 6 months ahead of everybody. So, speed is an advantage. And then two is you have to be able to get attention. I think Muse did this really well with their launch of their more most recent um personal agent. If you look at like their launch, like how much attention they got on that. exceptionally well. Um and then um yeah, distribution. Like can you get this out in into people? So, it’s almost like just a different weightings of things, a different type of founder that you need to to be

1:56:40 raising funds. But it’d be yeah, interesting to hear like this person’s perspective on that. And we have Xavier as well coming on the show on Friday to talk about this in startup program. So, I’m sure she’ll have some insights on that as well. >> Yeah. Right on. So, you just need to know Tuesdays and Fridays, Replit Live is coming at you. We got great content organized. Wednesdays, we get hands-on in office hours. And at the every at the end of every show, we do a little bit of dance off, you know, so >> a checkpoint, though? Do we >> Uh yeah, we can we can actually Ruth reminded folks here, so Hey everyone, uh

1:57:14 check in and hit the like button. So yeah, hit that like button on the video if you’re still watching. We appreciate that. Leave a comment on the video now or after it plays. Also super helpful. And then make sure you go over to community hub. If you go over to uh replit.com/community and you uh go to your profile. You’ll see that we have a check-in right now that’s open. So you can go ahead and check in. Uh also, I’m on a 49-day streak. Shoutout to 49-day streak. >> Yo, let’s go.

1:57:46 >> longest streak so far. So I think I just broke 47, so uh we’ve been cooking. We’d like to see what you’re cooking. Go ahead and check in on the background. We’re starting to think about creative ways to start rewarding the community for these types of engagements, so when we’re live, when we’re doing office hours, when we’re doing different things, there will be opportunities for you to check in and earn some community credits. And um yeah, stay tuned. There’s a lot more coming here, y’all. There’s a lot more coming. >> Guys, we’re getting a little lonely dancing up here by ourselves. Thank you, Kelly, who’s who’s posted a video of her

1:58:20 dancing with us, so won’t you post a little video of yourself dancing with us and post that for me. Tag Raimar, tag me. Love to see you get a little workout in, you know. >> I’m also saying a 39-day streak. Ruth is saying 88 days. Who’s got the highest streak? Let us know in the comments. >> I’m at 57. >> Oh, 57’s pretty good. I mean, that’s that’s pretty good. That’s almost 2 months of continuous >> I think what happens I have the boys hassling me for apps at home on the weekends, so I don’t miss the weekends. I’m building the weekends. [laughter]

1:58:54 >> All right, friends. Just at 8 days. 8 days is good enough. That’s just getting started, baby. You can just keep going. All right, everyone. >> weekends reclamation. Thank you. >> [music] >> Ooh, right at the drop. >> [music] [music]

1:59:25 >> Holy, 450 days. I don’t believe it. I need to see that posted on the internet. Hold up. Put the brakes on it. 450 days, I need to see that post. Go get your share card and put it up there. >> Let’s see the receipts. >> [music] >> All right, bye everyone. Catch you in office hours and later this week.