Kenhuang Academy
  • English

      • LangChain Tutorial

      Welcome to the LangChain tutorial. Based on official LangChain docs and the open-source langchain-ai/langchain project, this guide focuses on LangChain 1.0 agent development—from first run to production-ready apps.

      Table of Contents

      Basics

      1. Introduction — Ecosystem, 1.0 changes, LangGraph & LangSmith
      2. Installation — Python deps, API keys, virtual env
      3. Quick Start — create_agent and your first agent chat

      Architecture

      1. Architecture & Core Concepts — Agent loop, harness, LangChain vs LangGraph
      2. Models & Messages — Provider strings, messages, structured output

      Development

      1. Tools & Agents — @tool, tool calling, invoke / stream
      2. Middleware — Hooks, human-in-the-loop, context engineering
      3. Memory & Checkpoints — State, thread_id, persistence

      Applications

      1. RAG — Loaders, vector stores, retrievers, agentic RAG
      2. LangGraph Workflows — Graphs, nodes, workflows vs agents

      Operations

      1. Production & LangSmith — Tracing, eval, deployment
      2. Practical Examples — Q&A bot, multi-tool agent, orchestration
      3. Resources — Official docs and community

      Learning Path

      StageGoalChapters
      Day 1Run an agentIntro → Install → Quick Start
      Days 2–3Harness & toolsArchitecture → Models → Tools & Agents
      Week 1RAG & persistenceMiddleware → Memory → RAG
      Week 2Orchestration & prodLangGraph → LangSmith → Examples

      Prerequisites

      • Python 3.10+ basics
      • Familiarity with LLM / Chat APIs
      • At least one provider API key (OpenAI, Anthropic, etc.)