Learn how to develop advanced LLM applications using multiple specialized agents in this 56-minute technical talk. Explore the fundamentals of multi-agent frameworks and discover how LangGraph enables efficient task specialization across different domains. Master various multi-agent workflow patterns including router, consolidator, and sequential models while understanding the implementation of nodes, edges, and conditional logic. Follow along with hands-on exercises to create supervisor agents and see practical demonstrations of portfolio website creation and research report generation. Gain insights into optimizing multi-agent systems for real-world applications, with a focus on improving scalability and performance through distributed task management. Perfect for developers looking to overcome the limitations of single-agent systems and build more sophisticated LLM applications through collaborative agent architectures.
Multi-Agent Frameworks for LLM Applications: From Theory to Implementation
Overview
Syllabus
Introduction to Multi-Agent Frameworks
Why Do We Need Multiple Agents? Understanding Agentic Behavior
Types of Multi-Agent Workflows: Router, Consolidator, and Sequential Models
Building Multi-Agent Scenarios: Nodes, Edges, and Conditional Logic
Setting Up a Multi-Agent Workflow: A Step-by-Step Guide
Hands-On Exercise: Creating a Supervisor Agent
Practical Demo: Portfolio Website Creation and Research Report Generation
Q&A: Optimizing Multi-Agent Systems and Real-World Applications
Closing Remarks: Resources and Next Steps
Taught by
Data Science Dojo
Related Courses
-
Understanding AI Agents - From Basic Concepts to Multi-Agent Systems
-
Understanding Microsoft's Magnetic-One: A Generalist Multi-Agent AI System
-
Studying the Behavior of Generative AI-Based Agents in Multi-Agent Systems
-
Multi AI-Agents Reasoning with LLMs - Python Code Examples
-
Building Adaptive Multi-Agent AI Systems: From Theory to Implementation
-
Task-Optimized Multi-Agent Communication Protocols with G-Designer
