A hands-on technical session for engineers who want to see AI agents work with real data and run efficiently at scale. The first half treats a database like a game world your agent can explore and reason over - built on DocumentDB and Postgres.

The second half turns practical: container-enabled asyncio patterns for Python AI workflows, covering event loops, task queues, and non-blocking LLM calls. Attendees leave having both explored an agent-driven data workflow and refactored real async code live on screen.