- Home
- Courses
- LangChain & LangGraph for AI Agents
- Agent Foundations and LangChain Basics
Micro Free Course — 100% Free Learning
All lessons in this module are free to learn. Sign in with Google to save your progress.
Agent Foundations and LangChain Basics
What an AI agent actually is (vs chains, vs workflows); the LangChain ecosystem in 2026 (LCEL, Hub, Smith); building your first chains with LCEL; tools, toolkits, and tool calling.
Module Content
What an AI Agent Actually Is
A practical definition. Strip away the hype: agents are LLMs that decide what to do next. FIND_VIDEO: search 'what is an AI agent definition explained' — recommended channel: LangChain / AI Engineer / Harrison Chase. Aim for 10 min or under.
Recap — Agents vs Chains vs Workflows
The terminology is muddy. Here's the distinction that actually matters when you're building.
LangChain in 2026 — What's Changed, What's Stable
LangChain has evolved fast. Here's what's current, what's deprecated, and where to put your effort. FIND_VIDEO: search 'langchain 2026 LCEL latest version' — recommended channel: LangChain / Sam Witteveen / James Briggs. Aim for 11 min or under.
Recap — The LangChain Ecosystem (LCEL, Hub, Smith)
LangChain isn't just one library. The ecosystem has four pieces; you'll use all of them.
Building Your First Chain with LCEL
The pipe-based composition pattern. Once you see it, every LangChain example becomes readable. FIND_VIDEO: search 'LCEL LangChain Expression Language tutorial' — recommended channel: LangChain / Sam Witteveen. Aim for 10 min or under.
Recap — LCEL Patterns You'll Use Constantly
Five LCEL patterns cover 90% of chain-building. Master them; everything else is variations.
Tools, Toolkits, and Tool Calling in LangChain
The @tool decorator and bind_tools API. The foundation for agentic behavior in LangChain. FIND_VIDEO: search 'langchain tools tool calling tutorial' — recommended channel: LangChain / Sam Witteveen / Greg Kamradt. Aim for 11 min or under.
Recap — Designing Tools for Agent Use
Tools turn LLMs into action-takers. Design them well — the LLM decides which to call based on the descriptions.