Conclusion

Conclusion

In this lesson, you enhanced your agent’s capabilities by adding memory, structured output, and human-in-the-loop interaction. With all these, you can use the built-in features that come with LangGraph and LangChain or roll out your own solution. For example, instead of adding a breakpoint to the localizer app, you could have printed the advice from the AI agent and then let the human edit the final output from the formatter after the graph finished executing.

You’re nearing the completion of the course. In the final lesson, you’ll learn some techniques to help you debug your agents. This becomes more and more useful as the complexity of your agents increases.