Real-time voice-driven AI workspace assistant designed to make developer workflows completely hands-free using low-latency transcription and LLM reasoning.
Captures live microphone audio using SoundDevice and streams to Speechmatics WebSocket API with sub-second latency.
Interprets natural language commands using Gemini AI to run system processes, scripts, and manage files hands-free.
Integrates Pyttsx3 text-to-speech to provide clear verbal confirmations and execution statuses back to the developer.
# 1. Clone repository
git clone https://github.com/MuhammadAbubakar9/devvocal-agent.git
cd devvocal-agent
# 2. Set up virtual environment
python -m venv venv
venv\Scripts\activate
# 3. Launch Agent
python agent.py