Session Spotlight

Cameron Vetter

Camp Counselor

AutoGen: Empowering Next-Gen Multi-Agent LLM Apps

Event Logo

Tuesday, July 30, 2024 - 9:00 PM UTC, for 1 hour.

Regular, 60 minute presentation

Room: African 40

AI
LLM
AutoGen
Design Patterns
Python

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
favorited by:
Dave Boulineau Drew Nys Jason Cairol Joel Keyser Brett Allenstein Matt Frye Ross Larson Cameron Vetter Andy Bottom Adam Kerr Timothy Dooley