LangChain Agents: Build It From Scratch vs. create_agent

Playlist AI agent from Zero to Pro:    • AI agent tutorial from zero to pro   In this video we build an AI agent loop from scratch in plain Python, then rebuild the same agent with LangChain create_agent. Real-world example: an order support agent that looks up an order, calculates a refund and creates a ticket. What you will learn: What the agent loop is (Perceive, Reason, Act, Observe, Respond) How the message stack grows turn by turn Tool calling: the model asks, your code runs the tool Two ways a loop ends: task complete and the max_turns safeguard Build #1: the agent loop by hand with a for loop (no library) Build #2: the same agent with LangChain create_agent Why the same agent takes a different path for three different orders Chapters: 00:00 Intro [0:44] Recap: tools (previous video) and Why we need a loop [2:59] Real-world example: order support agent [8:00] Code with concept on The five-phase agent cycle, The message stack, Build #1: the loop by hand (code), Build #2: create_agent [44:00] Challenge and next video Course page: https://agenticcraft.aimentor360.com/... Code: [GitHub link] Next video: RAG and LangSmith tracing. #AIAgents #LangChain #Python