From Queries to Agents: The Next Era of Data Retrieval on PostgreSQL | POSETTE 2026

Why watch

This talk explores how PostgreSQL is evolving to support AI agents in production through emerging retrieval patterns. It covers the Model Context Protocol (MCP) for agent-database communication, common pitfalls when agents generate SQL blindly, and how blended retrieval combining vector similarity, relational SQL, and graph-aware traversal creates reliable data access. Designed for developers building AI-powered applications who need practical patterns for safe, context-aware database interactions with PostgreSQL.