Hello, and welcome to the Agentic AI with LangChain and LangGraph course.

You’ll begin by learning how to design agentic workflows using LangGraph, focusing on memory, iteration, and conditional logic. Through interactive lessons and labs, you’ll explore how LangGraph builds on LangChain to support adaptive decision-making using nodes, edges, and shared state.

As the course progresses, you’ll apply these skills to build self-improving agents using architectures like Reflection, Reflexion, and ReAct. You’ll also learn how to coordinate specialized agents in multi-agent systems, including agentic retrieval-augmented generation (RAG) pipelines that route queries to relevant data sources. Each concept is reinforced through hands-on labs designed to help you confidently implement advanced agentic AI systems.

This course is part of the IBM RAG and Agentic AI Professional Certificate , designed to provide you with the practical skills and knowledge to excel in developing advanced AI applications that leverage RAG, multimodal AI, and agentic AI systems.

Prerequisites

Python programming skills and experience are essential for this course, as you will immediately start building AI agents. Additionally, familiarity with core AI concepts and the LangChain framework is highly recommended. Here are some recommended courses if you are not familiar with the prerequisites:

• Python for Data Science, AI & Development

• Develop Generative AI Applications: Get Started

It is also recommended that you complete the previous courses in this Professional Certificate.

Objectives

After completing this course, you will be able to:

• Build agentic AI systems using LangChain and LangGraph to support memory, iteration, and conditional logic

• Design and implement self-improving agents using Reflection, Reflexion, and ReAct architectures

• Apply agent orchestration techniques to build collaborative multi-agent systems

• Implement agentic RAG systems that route queries and support retrieval-enhanced reasoning

Course outline

This course consists of three modules.

Module 1: Introduction to LangGraph

• Lesson 0: Welcome

• Lesson 1: Introduction to Agentic AI

• Lesson 2: LangGraph versus LangChain

• Lesson 3: Build a LangGraph Workflow

• Lesson 4: Module Summary and Evaluation

Key topics :

• Agentic AI fundamentals

• LangGraph workflows

• LangGraph nodes, edges, and state

Module 2: Build Self-Improving Agents with LangGraph

• Lesson 1: Build Reflection Agents

• Lesson 2: Advanced Self-Improvement with Reflexion Agents

• Lesson 3: ReAct: Integrating Reasoning and Action

• Lesson 4: Module Summary and Evaluation

Key topics :

• Self-improving agents

• Reflection and Reflexion patterns

• ReAct for reasoning and action

• Prompt engineering techniques

Module 3: Multi-Agent Systems and Agentic RAG with LangGraph

• Lesson 1: The Evolution from Single to Multiagent Systems

• Lesson 2: Build Multi-Agent Applications

• Lesson 3: Module Summary and Evaluation

• Lesson 4: Course Wrap-Up

Key topics :

• Multi-agent systems

• Agent orchestration

• Agentic RAG pipelines

• Query routing and retrieval

Tools/software

In this course, you will explore and utilize a variety of tools and platforms, including:

• LangGraph : To design and implement stateful agentic workflows that support memory, iteration, and conditional logic.

• LangChain : To integrate large language models (LLMs) with external tools and support core agent functionality.

• Large Language Models (LLMs) : Granite and GPT to build agents capable of reasoning, reflection, and interaction with external environments.

• Python : For scripting custom tools, defining workflows, and implementing agents and multi-agent systems.

Tips for success

• Stay organized and track key program milestones, risks, and stakeholder communications

• Consistently practice using the generative AI tools and techniques covered in the course

Congratulations on taking this step to advance your career by building expertise in agentic AI systems! Enjoy your learning journey.