About

Give your AI agent real utility. Go beyond prompts to run backend calculations, trigger workflows, and return structured outputs your app can trust. Learn to build custom tools with LangChain, enforce response schemas, and add branching logic for decision-making in composable, production-ready agents.

Key topics

1. Build a custom tool that performs backend logic using LangChain’s @tool decorator and Pydantic
2. Use with_structured_output() to enforce response schemas from your LLM
3. Introduce branching logic in your LangGraph agent with conditional edges like should_continue
4. Add decision flows based on the agent’s outputs