Development & Deployment

Running the Microservices Locally

Slothub relies on 4 distributed processes. In a real-world scenario, you would deploy these using container clusters such as AWS ECS (Fargate). For this workshop, we will start them sequentially on your local machine using the created .env files. Ensure you have the .env settings mentioned in the report.md.

1. Start the Backend API (Spring Boot)

The backend manages application logic and connects to the PostgreSQL database.

  • Open a terminal inside the backend/ folder.
  • Add your DB_HOST, JWT_SIGNER_KEY, email configurations, and AWS credentials to backend/.env.
  • Execute the application:
    # Using Maven wrapper
./mvnw spring-boot:run

The Spring Boot portal will be live on http://localhost:8080.

2. Start the AI Content Service (FastAPI)

The AI service manages AWS S3 and OpenAI API prompts.

  • Open a terminal inside the AI/ folder.
  • Verify OPENAI_API_KEY and AWS configs are loaded in AI/.env.
  • Create a virtual environment and start the Uvicorn server:
    uvicorn main:app --host 0.0.0.0 --port 8000

This service handles generative requests on http://localhost:8000/api/v1.

3. Start the LangGraph Agent-Core (Slozy)

For tutoring interactions and theory lookup via RAG.

  • Open a terminal in the agent-core/ folder.
  • Set LOCAL_DEV=1 for testing the memory locally rather than forcing an AWS Bedrock invocation.
  • Start the agent service. (Make sure you followed Python module requirements like pip install -r requirements.txt).
4. Start the Frontend (React Vite)

The unified user portal.

  • Navigate to the frontend/ folder.
  • Set VITE_API_BASE_URL to the Spring backend URL and VITE_FAST_API_BASE_URL to the Python API in frontend/.env.
  • Install packages and start the UI:
    npm install
npm run dev

Visit http://localhost:5173 locally. Log in, request materials from uploaded PDFs, and interact with the AI chat system to verify that the microservices are integrated perfectly!