Course Description
In this course, you'll learn to build a LangGraph-powered multi-agent system for software development and testing. Create developer and tester agents using the Critique–Revise pattern, connect them in a reflection loop, and visualize their workflows. Through hands-on coding and real-time debugging, you'll develop and evaluate agent performance. The course wraps up with fun mini-projects like FLAMES and Tic-Tac-Toe, helping you apply your skills to real use cases and automate development workflows efficiently.
Course curriculum
-
1
Course Introduction
- Course Kick off
- Introduction to the Course
-
2
Project Foundations
- Flow of the Project
- LangGraph Refresher
- Code Topology Cheat‑Sheet
- Environment Setup with LLM Initialization
-
3
Building Role-based Agents
- Agent Architecture: Critique - Revise Pattern Recording.mp4
- The Developer Agent Blueprint
- Developer Agent Design with Python REPL Tool Recording.mp4
- Testing Our Agentic System
- The Tester Agent Design
-
4
Building Multi-Agent System
- Building the Graph
- The Shared Graph State
- The LangGraph Flow Definition with Conditional Loop Logic.mp4
- Visualizing the Graph
- First Run – Demo Loop
- The Graph Execution
-
5
Mini Projects & Wrap-Up
- The FLAMES Game
- Hands-on: The FLAMES Game
- A Game of Tic-Tac-Toe
- Hands-on: The TIC-TAC-TOE game
- Key Benefits
Who Should Enroll
-
Developers & Testers looking to automate coding and QA workflows using AI agents.
-
AI/ML Enthusiasts eager to explore agentic systems with LangGraph and LangChain.
-
Tech Leads & Architects interested in building modular, looped LLM workflows for productivity.
-
Students & Builders who enjoy hands-on learning through real-world mini-projects like games.
What you will Learn
-
Hands-on experience building multi-agent systems using LangGraph with developer and tester roles.
-
Master the Critique–Revise loop to automate iterative code improvement and testing.
-
Learn to visualize and debug agent workflows using state tracking and Mermaid diagrams.
-
Apply concepts in real-world projects like FLAMES and Tic-Tac-Toe to solidify your skills.
About Instructor
Arun Prakash Asokan - Director Data Science, Leading Pharma Corporation

FAQ
-
Do I need prior experience with LangChain or LangGraph?
No prior experience is required. The course includes a refresher on LangGraph and guides you step-by-step through building the system.
-
What programming language is used in the course?
The course is built entirely in Python using google collab, making it accessible for most developers and ML practitioners.
-
Will I get access to the complete codebase?
Yes, you’ll have full access to the github repository with the code files for each section clearly mapped to the course videos.
-
Are there real-world applications beyond games?
Absolutely! The techniques you learn—like agentic critique-revise loops and automated testing—can be extended to software development, QA, and CI/CD workflows.
-
How long will the course take to complete?
You can complete the course in under2 to 2.5 hours, including both the walkthroughs and mini-projects.