AutoGen: Empowering Next-Gen Multi-Agent LLM Apps
Tuesday, July 30, 2024 - 9:00 PM UTC, for 1 hour.
Regular, 60 minute presentation
Room: African 40
Want to see two AI's play a game with each other? Or find out why it takes a team of AI to make a unicorn? Or see what seven AI's working together can accomplish? The future of AI applications lies in Multi-Agent LLM applications, and AutoGen is at the forefront of simplifying their implementation. I will introduce the concept of Multi-Agent LLM applications and talk about the design patterns. We will discuss why a Multi-Agent LLM is much more powerful than the single agent LLM's being created today. In this hands-on session, we will showcase the power of the Multi-Agent Conversation design pattern using OpenAI Models and the AutoGen framework. Through a live demonstration, we will illustrate how a team of three member agents and one manager agent collaboratively engage in a group conversation to efficiently solve a complex user task. Join us to learn about the next generation of AI applications, the patterns we will use to build them and see a live demonstration of them in action!
Prerequisites
Python experience will be helpful to understand the examples. A basic understanding of LLM will also be helpful.
Take Aways
- Large Language Models are great, but a group of them working together is even greater!
- Learn about the wisdom of the crowd, but the group is a group of LLM's
- Design pattern for LLM development of Next Generation Apps and a framework to make it easy
Supporting Resources
