ESN-PRISM
Edge-Native Recognition & Security Monitoring Intelligence
Edge AI
Computer Vision
Security
Python
Overview
Built an edge-native computer vision system that lets users query a live surveillance feed conversationally instead of reviewing footage by hand, pairing real-time recognition with a lightweight Llama reasoning layer and gesture-based control, engineered to run on a standard laptop with no dedicated GPU.
Key impact
Build complete and in testing; runs live on a Core i5 with no GPU.
Key features
- Query a live surveillance feed conversationally
- Real-time recognition and security monitoring
- Lightweight Llama reasoning layer
- Gesture-based control
- Runs on a standard laptop with no dedicated GPU