A voice-driven virtual companion on a Raspberry Pi 5 that listens, talks back, reacts with animated expressions and sees through its camera.
What it achieved
- Won MLH Best Use of ElevenLabs at HackUSF 2026
- Outperformed 80+ teams in a 24-hour hackathon
How I built it
- Event-driven Tkinter GUI with a finite-state machine coordinating real-time voice and animated facial states
- Agentic tool-calling pipeline using the OpenAI API with structured JSON schemas
- Real-time voice with ElevenLabs and image capture through the Pi Camera