Agenda — LangGraph¶
Session length: 3 hours | Difficulty: Advanced | Prerequisites: Tool-Calling Agents, Python dataclasses, TypedDict
What you will build today¶
A stateful multi-agent research pipeline: a planner node that decomposes a question, parallel researcher nodes that gather data, and a writer node that synthesizes the final report — with conditional routing based on quality checks.
Schedule¶
| Time | Topic | File |
|---|---|---|
| 0:00–0:20 | What LangGraph is and why it exists | 01-langgraph-overview |
| 0:20–0:50 | Nodes, edges, and the StateGraph API | 02-nodes-and-edges |
| 0:50–1:20 | Stateful graphs: TypedDict state, reducers | 03-stateful-graphs |
| 1:20–1:50 | Conditional routing: edges based on state | 04-conditional-routing |
| 1:50–2:15 | Multi-agent orchestration | 05-multi-agent-orchestration |
| 2:15–2:45 | Practice exercises | 06-practice-exercises |
| 2:45–3:00 | Interview questions review | 07-interview-questions |