Building Multi-Agent Applications: The good, the bad, and the downright crazy

Please RSVP via the Data Science London meetup: https://www.meetup.com/data-science-london/events/303482035/?slug=data-science-london&isFirstPublish=true

During this session, Sandy goes through various approaches he's seen and implemented for multi-agent LLM-based production applications. He'll show how we should be managing access to applications and data sources, as well as the importance of a strong user-based security model.
Utilising LangGraph, Azure OpenAI, various Azure Data Stores and ChainLit for a UI, expect multiple demos and fast-paced action. But staying true to how we can utilise this technology together to build, strong, more secure applications that have meaning to users, rather than just getting it done.
We'll end with a bit of fun, having a multi-agent workflow research various topics and execute a debate, with a supervisor giving a winner for the argument. Finally, settle the discussion of Star Wars vs Star Trek anyone?