The Open Source Approach: Building Production-Ready AI Apps with Postgres
Presented by:
Mike Josephson
Mike Josephson (MJ) is passionate about technology and the transformative effect it can have on his customers. He works tirelessly to remove customer roadblocks and ensure the success of those that he works with. For afterall it is in the success of ones customers that we are likewise successful. When he is not working you can find him spending time with his family or hiking the mountains near his house.
No video of the event yet, sorry!
It’s no secret that Postgres is uniquely positioned as being an ideal foundation for agentic AI applications. However, there are many different components you need to consider to ensure your application runs as expected in an enterprise environment: in a security hardened manner, grounded in verifiable information, with efficient token usage.
In this talk, we will show you how to leverage the open source tools (licensed under the PostgreSQL license) within our pgEdge Agentic AI Toolkit to address these common concerns and ensure you’re prepared for different architectural use cases.
We’ll cover at a high level:
The pgEdge MCP Server for PostgreSQL, a highly performant and fully featured MCP server for access to any standard PostgreSQL database, whether greenfield or pre-existing
Natural Language Agents for querying data, available via a command-line interface (CLI) or a web user interface (full-featured MCP clients written in Go, with Anthropic prompt caching for a 90% cost reduction)
pgEdge-vectorizer, a Postgres extension that automatically chunks text content and generates vector embeddings using background workers
pgEdge RAG Server, a dedicated API Server for performing Retrieval-Augmented Generation (RAG) of text based on content from a PostgreSQL database using pgvector
pgEdge-docloader, a utility that makes it easy to bring initial material online
VectorChord-bm25, a Postgres extension implementing BM25 ranked search for hybrid semantic and full-text searching
By the end of the session, you will have insights on how to safely deploy your AI applications to production while ensuring your data sovereignty, privacy, integrity, security, and performance is rock solid.
- Date:
- Duration:
- 50 min
- Room:
- Conference:
- Postgres Conference: 2026
- Language:
- Track:
- Dev
- Difficulty:
- Easy