Microsoft Developer hosts Abe Omorogbe showing how to build AI-powered agents on Azure backed by PostgreSQL, including secure connectivity, natural-language querying/analytics, and vector search, using Microsoft Foundry.

Build smart and secure agents with PostgreSQL on Azure using Microsoft Foundry

Join Abe Omorogbe to learn how to build AI-powered agents backed by PostgreSQL on Azure using Microsoft Foundry.

What this video covers

• Support for PostgreSQL on Azure in Microsoft Foundry
• How to securely connect agents to PostgreSQL data
• Using natural language for querying and analytics (natural language → SQL)
• Adding vector search to get more relevant results
• Running on Azure infrastructure powered by AMD processors for performance and efficiency at scale

Demo components mentioned

• PostgreSQL extension in VS Code
• A demo agent built in Microsoft Foundry
• A PostgreSQL MCP Server (Model Context Protocol)

Two use cases demonstrated

1. Natural language and SQL

   • Use natural language to query PostgreSQL data and perform analytics.

2. Incorporating vector search

   • Add vector search to improve relevance of results when interacting with PostgreSQL-backed data.

Timeline

• 0:00 Introduction
• 1:00 Support for PostgreSQL on Azure to Microsoft Foundry
• 1:30 PostgreSQL extension in VS Code
• 2:16 Demo agent in Microsoft Foundry
• 3:06 PostgreSQL MCP Server
• 3:58 Two use cases for using PostgreSQL for talking to your data
• 4:05 First use case: Natural language and SQL
• 6:25 Second use case: How to incorporate vector search
• 7:29 Wrap up

Links

• Watch the series: https://www.aka.ms/PostgreLikeAPro
• Guided demo (build your own agentic app): https://www.aka.ms/AzurePostgres-AI