Particle Autonomous AI Assistant

Particle is a personal AI chief of staff designed for 24/7 autonomous operation.

Installation

1. Install Python 3.11+ and create a virtual environment.
2. Install dependencies:

   pip install -r requirements.txt

3. Copy environment template:

   cp .env.example .env

Configuration

• Edit config.yaml for runtime settings (paths, polling, module options).
• Fill all required keys in .env:
  • GEMINI_API_KEY
  • OPENROUTER_API_KEY
  • TELEGRAM_BOT_TOKEN
  • TELEGRAM_HOME_ID
  • EMAIL_ADDRESS
  • EMAIL_PASSWORD
  • GOOGLE_CALENDAR_CREDENTIALS

Persona Cloning (Free Cloud Setup)

What it does

• Attends meetings as you
• Speaks in your cloned voice via ElevenLabs
• Shows your face via Deep-Live-Cam (local)
• Thinks and responds using Gemini/OpenRouter LLM

All free — setup steps

a. Hugging Face (STT):

• Sign up at huggingface.co (free)
• Get token from Settings → Access Tokens
• Add to .env as HF_TOKEN

b. ElevenLabs (voice clone):

• Sign up at elevenlabs.io (free tier = 10k chars/month)
• Go to Voice Lab → Add Voice → clone your voice (needs 1 min of audio)
• Copy the Voice ID
• Add to .env as ELEVENLABS_API_KEY and ELEVENLABS_VOICE_ID

c. Face clone (local):

• Save a clear photo as assets/my_face.jpg
• git clone https://github.com/hacksider/Deep-Live-Cam
• cd Deep-Live-Cam && pip install -r requirements.txt

d. Enable in config:

• Set clone.enabled: true in config.yaml
• For GPU: set clone.execution_provider: cuda

Pipeline diagram

Meeting audio → HF Voxtral STT → Gemini LLM → ElevenLabs TTS (your voice) + Deep-Live-Cam (your face) → back into the meeting

Usage

Run bootstrap validation:

python main.py

This loads config.yaml and .env, logs resolved settings with secrets masked, and exits.

Deployment (Docker)

1. Build image:

   docker build -t particle-agent:latest .

2. Start service:

   docker compose up -d

Deployment (systemd on Ubuntu 22.04)

1. Place project at /home/ubuntu/particle-agent.
2. Ensure virtual environment exists at /home/ubuntu/particle-agent/.venv.
3. Install service:

   sudo cp particle.service /etc/systemd/system/particle.service
   sudo systemctl daemon-reload
   sudo systemctl enable --now particle.service

4. Inspect logs:

   journalctl -u particle.service -f

Project Layout

particle-agent/
├── main.py
├── cli.py
├── orchestrator.py
├── config.yaml
├── .env.example
├── requirements.txt
├── docker-compose.yml
├── particle.service
├── modules/
├── context/
├── data/
└── logs/