Abstract: The dawn of AI agents is upon us, with tech visionaries like Bill Gates, Andrew Ng, and Vinod Khosla highlighting their monumental potential. Recent breakthroughs in large language models (LLMs( have opened unprecedented possibilities. They promise to revolutionize nearly every industry.
While chatbots like ChatGPT, Claude, and Gemini have shown remarkable potential, the future lies in generative AI agents capable of executing complex tasks for users. These agents can conduct detailed research, automate project management, analyze and synthesize information from vast data sources, and optimize business processes across various domains.
By completing this workshop, you will gain a high-level understanding of AI agents. This will include their workflows, types, and use cases, along with an introduction to LangGraph and its relationship to LangChain. Lesson 1 covers the fundamentals of AI agents, exploring their various applications and capabilities. You will learn to install LangGraph and create a simple AI agent.
In Lesson 2, you will engage in a hands-on demo to develop a multi-agent app for researching topics. This session includes conceptualizing your application, applying it to a graph, and evaluating LLMs and tools. Lesson 3 advances your skills by using LangGraph to create an agent leveraging Retrieval Augmented Generation (RAG), which will include access to the ChromaDB vector database.
Session Outline:
Lesson 1: Understanding AI Agents and LangGraph
Get a high-level understanding of how AI agents work, such as the workflows, types, and use cases. There will be an introduction to LangGraph, its main capabilities, and its relationship to LangChain. You will then see how to install LangGraph and create a simple AI agent.
Lesson 2: Create a Multi-Agent App
This will be a hands-on demo of creating a multi-agent app that will be used to research a topic. You will first learn how to sketch out the concept and apply it to a graph. You will then see how to evaluate an LLM and the related tools.
Lesson 3: Using RAG with Your AI Agents
Let’s take LangGraph to the next level. We’ll see how to use this powerful tool to create an agent that leverages RAG (Retrieval Augmented Generation). This will involve accessing an external vector database.
Bio: Tom (@ttaulli) is an author and IT consultant, specializing in AI and its applications in programming. He has written several books, including AI-Assisted Programming: Better Planning, Coding, Testing, and Deployment (O’Reilly Media). In his consulting role, Tom advises companies on AI strategies and implementations. He has also taught IT courses at UCLA, PluralSight, and O'Reilly Media, providing lessons on using Python to develop deep learning and machine learning models. Throughout his career, Tom has founded several tech startups, including BizEquity, ExamWeb, and Hypermart.net, the latter of which was sold to Infospace.

Tom Taulli
Title
Author and Consultant, Advisory Board Member | TadHealth


