Hackathon Mockup Backend

Simple FastAPI backend for video/audio analysis with Whisper and Gemini.

Project Layout

• main.py - current FastAPI entrypoint kept for uvicorn main:app
• app/main.py - package entrypoint alias for a cleaner import path
• data/keywords.json - keyword dictionary used by detection
• temp/ - runtime scratch space for extracted media

Dependencies

• FastAPI - pip install fastapi
• Uvicorn - pip install uvicorn
• Whisper - pip install openai-whisper
• Multipart form support - pip install python-multipart
• Gemini SDK - pip install google-genai
• Dotenv support - pip install python-dotenv
• FFmpeg - sudo apt install ffmpeg

Run Locally

1. Create and activate a virtual environment:

   python3 -m venv .venv
   source .venv/bin/activate

2. Install Python dependencies:

   pip install -r requirements.txt

3. Create your local environment file:

   cp .env.example .env

4. Open .env and set your GEMINI_API_KEY value.
5. Make sure ffmpeg is installed on your machine.
6. Start the FastAPI server:

   uvicorn main:app --reload --port 9000

7. Open the API at:

   http://127.0.0.1:9000
   

Run With Docker

1. Build the image:

   docker build -t hackathon-mockup-backend .

2. Run the container and pass your Gemini API key:

   docker run --rm -p 9000:9000 --env-file .env hackathon-mockup-backend

3. Open the API at:

   http://127.0.0.1:9000
   

Run With Docker Compose

1. Start with compose:

   docker compose up --build

2. Run in detached mode:

   docker compose up --build -d

3. Stop services:

   docker compose down

Notes

• The root keywords.json is kept as a compatibility link to data/keywords.json so the existing entrypoint keeps working.
• No application logic was refactored as part of this cleanup.