Borderless Lakehouse, Smart Storage & Knowledge Catalog
Read the Portuguese translation →
Author: Google Cloud · Watch on YouTube
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: Borderless Lakehouse, Smart Storage & Knowledge Catalog · Channel: Google Cloud · 01:34 · Watch on YouTube
Topics: AI agents, Data, databases and embeddings, Privacy and data protection
Summary
Section titled “Summary”Gemini’s Agentyc Data Cloud brings grounded AI intelligence to business data where it already resides, avoiding the cost and delays of migrating it into a central silo. Its three capabilities are Knowledge Catalog, Borderless Lakehouse, and Smart Storage, which provide business context, enable zero-copy queries, and enrich unstructured data in place. The approach keeps data within existing security and governance controls while reducing data-engineering toil. Gemini connects to data stores with zero-copy execution and code transparency, helping teams move from data discovery to predictive ML and business activation.
Key points
Section titled “Key points”- AI agents can give inaccurate answers when they lack company-specific context for metrics such as churn or net margin.
- The Agentyc Data Cloud brings grounded intelligence to existing data stores without requiring a central data migration.
- Knowledge Catalog maps schemas, lineage, and business definitions without re-indexing.
- Business metrics can be defined once and used consistently by Gemini agents.
- Borderless Lakehouse runs zero-copy queries on data in place, with no variable egress fees.
- Keeping data in place helps it remain within existing security boundaries.
- Smart Storage enriches unstructured objects such as images, PDFs, and physical scans in place.
- Pre-built data cloud skills let engineers and developers describe desired outcomes to Gemini.
- Gemini connects directly to data stores with zero-copy execution and full code transparency.
Chapters
Section titled “Chapters”| Start | Chapter | Summary |
|---|---|---|
| 00:00 | The challenge of grounding AI in enterprise data | The video explains why agents need company-specific context and how traditional data migration creates cost and delays. |
| 00:31 | Three data capabilities | Knowledge Catalog provides business context, Borderless Lakehouse queries data in place, and Smart Storage enriches unstructured data. |
| 01:04 | From data work to business activation | Gemini connects to data stores with zero-copy execution and code transparency, helping teams move toward predictive ML and business activation. |
Quotes
Section titled “Quotes”reasoning without context is just a guess. — 00:00
You define a metric like churn or net margin once and then every Gemini agent uses that exact definition. — 00:31
Gemini runs zero-copy queries with zero variable egress fees. — 00:31
Smart storage enriches those objects in place and writes context directly back into the storage object — 01:04
Gemini connects directly to your data stores with zero-copy execution and full code transparency. — 01:34
Concepts and entities
Section titled “Concepts and entities”Agentyc Data Cloud, Gemini, Knowledge Catalog, Borderless Lakehouse, Smart Storage, zero-copy queries
Full transcript (original language)
Section titled “Full transcript (original language)”00:00 Point AI agent to an enterprise database through an API and it hits an accuracy wall fast. Not because the model can’t reason, but because reasoning without context is just a guess. It doesn’t actually know what your specific company means by metrics like net margin or churn. Traditionally, fixing that meant migrating petabytes of data into a central silo and building endless extraction and cleaning pipelines, incurring massive expenses and multi-quarter delays. Gemini bypasses that migration trap entirely with the Agentyc Data Cloud, bringing grounded intelligence directly to where your
00:31 business data already lives, inside the security and governance controls you already trust. It does this with three core data capabilities. First, knowledge catalog gives every agent your business context. It maps your schemas, lineage, and business definitions with no re-indexing required. You define a metric like churn or net margin once and then every Gemini agent uses that exact definition. Second, borderless lakehouse queries your data in place. Gemini runs zero-copy queries with zero variable egress fees. Because your data never
01:04 moves into a new silo, it never leaves your security boundary. And third, smart storage unlocks the unstructured dark data your warehouses can’t see. Up to 90% of enterprise data is trapped in images, PDFs, and physical scans. Smart storage enriches those objects in place and writes context directly back into the storage object, letting agents inspect physical work directly. For data engineers and developers, this eliminates months of toil. Using pre-built data cloud skills in Gemini, you simply describe the outcome you
01:34 want. Gemini connects directly to your data stores with zero-copy execution and full code transparency. By uniting knowledge catalog metadata, the zero-copy borderless lakehouse, and smart storage under one govern control plane, Gemini takes teams from data discovery to predictive ML and business activation in minutes.