Automated Calling
A production-ready, end-to-end fully-local AI voice agent that runs on consumer hardware without cloud APIs or expensive subscriptions.
Quick Links
- 📖 Quick Start - Get running in 5 minutes
- 🏗️ Architecture - System design overview
- ⚙️ Configuration - Setup guide
- 🧠 Models - Component deep dives
- 🔌 n8n Integration - Webhook setup
- 🔧 Engineering - Technical details
- ❓ Troubleshooting - Common issues
Features
✅ Full Local Execution - No cloud APIs, no data sharing
✅ CPU Optimized - Tuned for i3-9100F + GT 730 (4GB VRAM)
✅ Low Latency - Sub-second response times
✅ n8n Compatible - Webhook integration built-in
✅ Production Ready - Error handling, logging, state management
Performance
Conservative Path (Default)
- Response Time: 1.0-1.2s per turn
- VRAM Required: ~1.5GB
- Status: ✅ Production-Ready
Modern Path (Advanced)
- Response Time: 0.35-0.50s per turn (2.5-3x faster)
- VRAM Required: ~3.6GB / 4GB
- Status: ✅ Stable, GPU-optimized
Hardware
Optimized For: i3-9100F + GT 730 (4GB dedicated VRAM)
Minimum:
- Quad-core CPU (3.5+ GHz)
- 8GB RAM
- 2GB dedicated VRAM (Conservative Path)
- Microphone + Speakers
Recommended:
- 6+ core CPU
- 16GB RAM
- 4GB dedicated VRAM
- USB headset