tubeshowcase●1,673,666 videos♡ My collection 0
← Learn to Code

Learn to Code13:57

Embeddings, AI, and SQL Server 2025 with David Mauri, in Mr. Maeda's Cozy AI Kitchen

Microsoft Developer ↗

From the creator

In this episode of the Cozy AI Kitchen, we’re cooking up something special. Join host John Maeda and special guest Chef David Mauri, PM for AI and SQL at Microsoft, as they explore how SQL Server 2025 and Azure SQL are evolving with AI-powered capabilities like vector search, embeddings, and LLM integration. From generating embeddings with T-SQL to building full RAG (Retrieval-Augmented Generation) patterns and exposing SQL as REST endpoints, this episode is packed with practical demos and delicious metaphors. Whether you're a seasoned data chef or just getting started with AI in databases, this one’s for you. Chapters: 00:00 - Intro: SQL is like bread 🍞 00:21 - Meet Chef David Mauri 01:04 - Using AI to find SQL samples 01:28 - Embeddings in SQL Server 2025 02:03 - Bring your own model (BYOM) 02:42 - Generating embeddings with T-SQL 03:22 - Storing and chunking embeddings 04:41 - Chunking strategy explained 05:24 - Vector similarity search demo 06:34 - Searching across multiple fields 07:57 - Calling GPT-4 from SQL 09:00 - Building an agent pattern in SQL 10:01 - REST endpoints with Data API Builder 11:04 - Full-stack AI app with SQL 12:22 - Final thoughts and call to action What You'll Learn: How to generate and store embeddings in SQL Server 2025 How to perform vector similarity search using T-SQL How to integrate LLMs like GPT-4 into your SQL workflows How to expose SQL logic as REST or GraphQL endpoints How to build secure, scalable AI apps with your own data 📂 Links

Published 2025-07-29 · Watch on YouTube ↗
Availability and playback restrictions are controlled by the original channel.

Keep exploring Learn to Code