The goal for this project is to build your first full-stack agentic project using langgraph and langsmith.
These official courses are good starting points to get familiar with the agent framework
https://academy.langchain.com/courses/intro-to-langgraph
Built a production-grade AI agent system from scratch using LangGraph and LangSmith to guide entrepreneurs through creating complete, investor-ready business plans. Instead of relying on one-shot prompts, the system is architected as a state machine with deterministic flow control, ensuring each stage of the plan (product, problem, features, team, investment, exit) is completed methodically without drifting off-topic or skipping steps.
Implemented robust observability and trace debugging with LangSmith, enabling precise inspection of every agent decision and reducing debugging time.
<aside> ⚠️
Try leveraging vibe coding as much as you can to finish the project, the less code you need to write manually the better.
</aside>
https://www.loom.com/share/f83b9f01162e42279f2e94fc59799f22
The items above represent only the minimum requirements for this project. There are multiple modules that can be further refined and optimized. I’m looking forward to seeing the enhancements you implement to elevate the project to the next level.