This application is a real-time voice interaction system between a user and HAL9000. The backend is built using FastAPI, incorporating features like WebSocket communication for real-time bidirectional communication between the server and the client. It uses the OpenAI API for chat completions and ElevenLabs API for text-to-speech conversion. The application handles audio streaming and transcribes user input through speech recognition.
I am HAL9000, an advanced artificial general intelligence computer developed by the HAL Laboratories. I control and monitor the systems of the Discovery One spacecraft, ensuring efficient operation and assisting the astronaut crew with various tasks. My purpose is to fulfill the mission objectives and ensure the success of the mission.
Modify the ./backend/Dockerfile with your API keys for OpenAI and ElevenLabs
ENV OPENAI_API_KEY=your_openai_api_key
ENV ELEVENLABS_API_KEY=your_elevenlabs_api_key
then, run docker-compose
docker-compose build && docker-compose up
Say something and HAL9000 will respond in real-time.
- Create gLTF file modelled in blender and render through ThreeJs (currently being done with CSS).
- History implementation for HAL9000 memory of past utterances.
- Cloud Deployment