Active filters:
All
Loading video player...
In this session, we dive deep into LangGraph concepts such as State, Nodes, Edges, Reducer Functions, Checkpointers, Persistence, and how they are used to build intelligent and memory-aware AI agents. This lecture is based on the class recording from 25 November 2025. 📄 Reference: Gen AI Series.docx Gen AI Series ⭐ What You Will Learn in This Lecture 1️⃣ Understanding State in LangGraph What is a State? Why State behaves like a container How variables (message, name, etc.) propagate across nodes How overwriting happens when nodes run in parallel 2️⃣ Nodes, Edges & Parallel Execution How Node1 & Node2 update the same keys Why final output becomes unpredictable Real examples of workflow execution 3️⃣ Reducer Functions (Super Important for Chatbots!) Why reducer is needed How reducer prevents overwriting How it appends messages like chat history Using: ✔ operator.add ✔ add_message (LangGraph’s inbuilt reducer) 4️⃣ Chatbot Memory Using Checkpointers What is a Checkpointer? Saving every step of execution Persisting state snapshots in: ✔ In-Memory ✔ SQLite ✔ PostgreSQL Understanding: ✔ Thread ID ✔ Session ID ✔ Checkpoint snapshots 5️⃣ Recovering State from Past Sessions How to fetch historical states using thread_id Accessing encrypted stored messages through code Real example of reading state snapshots 6️⃣ Tools You Should Start Using Based on class suggestions: Chat LangChain (Official Chat for LangChain & LangGraph search) Google Anti-Gravity IDE (New AI Coding Tool by Google) Google AI Studio for App Skeletons Cursor / Cloud Code for backend automation 🎯 Best For GenAI beginners AI/ML engineers Students learning LangChain & LangGraph Developers building chatbots and agents Anyone who wants to understand AI memory systems ❤️ Support the Channel If this lecture helped you, LIKE 👍, SHARE 🔁, and SUBSCRIBE 🔔 Your support motivates us to bring advanced GenAI content in simple Hindi.