Python chatbot project with:

• FastAPI/LangServe backend
• Gradio frontend
• External LLM API integration (bring your own API key)

docker build -f composer/Dockerfile -t chatbot-app .

docker run --rm -p 7860:7860 -e NVIDIA_NIM_API_KEY=your_key chatbot-app

Then open:

• http://localhost:7860

NVIDIA_NIM_API_KEY is supported and automatically mapped to NVIDIA_API_KEY inside the container.

NVIDIA_NIM_API_KEY=your_key docker compose -f composer/docker-compose.yml up --build

Create .env in the project root:

NVIDIA_NIM_API_KEY=your_key

Then run:

docker compose -f composer/docker-compose.yml up --build

Open:

• http://localhost:7860

• Frontend runs on port 7860 (container and host mapping).
• Backend runs internally on port 9012.
• Logs stream to your terminal (docker logs / docker compose logs -f).
• Stop cleanly with Ctrl+C (Compose) or docker stop <container_id>.

• config_LangGraph/README.md
• PyKGML/README.md
• chatbot/README_docker.md (prebuilt image packaging and end-user run guide)

This project development was supported by Prof. Licheng Liu (University of Wisconsin, Madison; ECAI Lab), Prof. David Mulla (The University of Minnesota, Twin Cities; AI-LEAF Institute), and Prof. Ce Yang (The University of Minnesota, Twin Cities; Agricultural Robotics Lab).