In partnership with

What’s in today’s newsletter:

Snowflake integrates with Oracle Cloud for AI advancement 🤝

Replit acquires Chroma to boost AI memory continuity 🤖

BigQuery queries Databricks, Snowflake data seamlessly 🌟

Free Virtual Summit: How Apple, NVIDIA & Microsoft AI Build for Data + AI


What changes when your next database user is an AI agent?

Join engineers from Apple, NVIDIA, Microsoft AI, Coinbase and AWS at NEXT QUERY 2026— a FREE online Data + AI Infrastructure Summit on November 4.

Explore 20+ sessions on real-time analytics, AI agent context, data governance, query performance and infrastructure costs. Get practical lessons from the teams building and running these systems in production.

November 4 · Starts at 9 AM PT / 12 PM ET
Free to attend. Register to receive the recordings, even if you can’t join live.


Register for Free

Also, check out the weekly Deep Dive - Fabric and SQL Server

ORACLE-SNOWFLAKE PARTNERSHIP

Courtesy: Snowflake

TL;DR: Snowflake and Oracle partnered to integrate Snowflake natively on OCI, enhancing multi-cloud AI capabilities, data sharing, and operational agility, enabling enterprises to accelerate AI projects and foster innovation efficiently.

  • Snowflake partners with Oracle to integrate its data platform natively on Oracle Cloud Infrastructure (OCI).

  • The collaboration enhances multi-cloud performance, scalability, security, and seamless data sharing capabilities.

  • Enterprises can accelerate AI projects by combining OCI's AI services with Snowflake's advanced data processing.

  • This alliance simplifies data management, improves operational agility, and fosters innovation for AI-driven businesses.

Why this matters: The Snowflake-Oracle partnership streamlines multi-cloud data integration, enabling faster AI deployment and better data governance. This collaboration reduces complexity and costs, empowering businesses to unlock new insights and maintain a competitive edge through scalable, secure, and innovative AI-driven cloud solutions.

VECTOR DATABASE

TL;DR: Replit acquired Chroma to integrate AI memory for better context retention, allowing cloud shutdown in 60 days while preserving AI state, boosting developer productivity, privacy, and reducing operational costs.

  • Replit acquired Chroma to integrate advanced AI memory, enhancing context understanding in coding interactions.

  • Chroma’s technology enables AI agents to retain memory across sessions, improving long-term user engagement.

  • Users can shut down cloud environments in 60 days while maintaining AI memory continuity for efficiency and privacy.

  • This move aims to create smarter coding assistants, boosting developer productivity and reducing operational costs.

Why this matters: Replit's acquisition of Chroma pioneers AI memory retention across sessions, enabling smarter coding assistants that enhance developer productivity. Maintaining AI memory while allowing cloud shutdowns balances efficiency and privacy, potentially lowering costs and transforming long-term AI interactions in software development.

BIGQUERY/GCP

TL;DR: Google Cloud’s BigQuery now queries Databricks and Snowflake data directly, enabling multi-cloud analytics without data movement, boosting flexibility, reducing vendor lock-in, and enhancing BigQuery’s competitiveness in hybrid environments.

  • Google Cloud's BigQuery now supports querying data directly from Databricks and Snowflake platforms.

  • This integration enables federated queries, eliminating the need to move or duplicate data across clouds.

  • The feature is initially launching in public preview to gather user feedback and improvements.

  • Cross-platform querying enhances flexibility and reduces vendor lock-in for multi-cloud data analytics.

Why this matters: Google Cloud’s new BigQuery interoperability with Databricks and Snowflake enables seamless multi-cloud data analysis, cutting costs and complexity by eliminating data replication. This fosters flexibility, breaks vendor lock-in, and strengthens Google’s position as a versatile leader in hybrid cloud analytics environments.

EVERYTHING ELSE IN CLOUD DATABASES

DEEP DIVE

I seem to be at Microsoft HQ on Bay Street so much, I should ask building security for a security pass.

Anyhow, the Fabric/SQL Server presentation was a combination of points from two conferences (FabCon and SQLCon Europe).

I should state that lately there has been a dearth of Fabric news here, but I wish to rectify that today, and going forward. As a matter of fact, I have many Azure readers of the newsletter, possibly the largest cohort.

Based on the presentation slides from the Microsoft event (see below), the core message Microsoft is telegraphing is a fundamental repositioning of their data ecosystem: SQL Server and Azure SQL are no longer just relational databases; they are being engineered as the native, hybrid memory layer for agentic AI and enterprise Copilots.

Here is a deep-dive breakdown of the strategic narratives and architectural patterns Microsoft was conveying to attendees:

1. The Death of the Standalone Vector Database

Microsoft is aggressively pushing the narrative that enterprises do not need specialized, siloed vector databases.

  • They are positioning Azure SQL as a unified "Operational + Vector Database" capable of handling Retrieval-Augmented Generation (RAG) to lower cost and architectural complexity.

  • The slides highlight new capabilities like native "Vector Index & Search" rolling out across the versionless SQL engine.

  • Beyond simple search, they are marketing SQL as the "Enterprise Memory" store, where AI agents can securely store context, knowledge, and decisions with continuity using the Microsoft Agent Framework.

  • They are even repurposing legacy features for the AI era, such as using SQL's ledger capabilities to create "tamper-evident chat history" for AI interactions.

2. Infrastructure Built for "Agentic" Users

A major theme is that databases must now serve two masters: human developers and autonomous AI agents.

  • Microsoft is deploying "Agent skills for the whole data journey," offering 70 unique skills across 5 plugins.

  • These AI agent skills are deeply embedded into tools like SQL Server Management Studio (SSMS) for diagnostics and recovery, Visual Studio Code for app development, and the Fabric Database Hub for estate inventory.

  • To enable frictionless agent interaction, Microsoft highlighted the "SQL MCP Server" (Model Context Protocol), providing a standardized REST API pathway for AI agents to interface directly with SQL.

3. The "One Estate" Hybrid Reality

Microsoft acknowledges that the enterprise migration to the cloud is not a sudden cliff, but a prolonged hybrid reality. Instead of forcing a hard migration, they are trying to own the control plane.

  • The "SQL hybrid architecture" slide illustrates a unified estate spanning on-premises data centers (running the upcoming SQL Server 2025 for local data residency) and the Azure SQL cloud estate (Hyperscale and Managed Instances).

  • The connective tissue for this hybrid world is Azure Arc for unified inventory, patching, and governance, alongside Microsoft Entra ID for unified identity management.

  • For workloads ready to scale, they heavily pushed Azure SQL Database Hyperscale, which now expands up to 128 TB of storage and 192 vCore options.

4. Fabric as the Connective Enterprise Hub

Microsoft Fabric is positioned as the central nervous system connecting user productivity to enterprise data.

  • The "Fabric IQ in Microsoft Copilot Chat" architecture slide details how user prompts in the Microsoft 365 boundary are grounded securely.

  • The architecture queries Data from Power BI semantic models within the Fabric tenant, utilizing Microsoft Graph for context, and sends modified prompts to a secure Azure OpenAI instance where customer data is not used to train foundational models.

  • The integration is tightening, with "SQL database in Fabric" introducing workspace-level private links and native migration experiences directly within Fabric.

5. Aggressive Ecosystem Lock-In and Skill Building

To ensure adoption of this new architecture, Microsoft is heavily lowering the barrier to entry while formalizing the required human skill sets.

  • The Freemium Hook: They are offering a highly generous free tier for Azure SQL Database, giving users up to 10 databases, 100,000 vCore seconds of compute, and 32 GB of storage per month for the lifetime of a subscription.

  • Structured On-Ramps: Their call-to-action funnel provides clear next steps, including a 60-day free Fabric trial and free infrastructure assessments.

  • Formalizing the Role: Microsoft is pushing a new "SQL Database AI Developer Certification" to train engineers on building LLM-integrated apps, AI-augmented SQL, and responsible AI governance.

The Bottom Line: Microsoft wants organizations to view Azure SQL and Fabric not just as places to store data, but as the foundational infrastructure required to safely and securely deploy autonomous AI agents across a hybrid enterprise footprint.

Note that I am a Senior Data Architect/Cloud Database newsletter editor and not a professional photographer.

Enjoy your Monday!

Gladstone Benjamin

Keep Reading

View more
caret-right
© 2026 Cloud Database Insider.
beehiivPowered by beehiiv