GodView AI
Real-time clinical anomaly detection with computer vision and automated alerts
A real-time patient-monitoring platform that combines YOLO-based visual inference, role-specific access, persistent clinical data, and automated alert delivery in one operational workflow.
The 15-second brief
At a glance
- Outcome
- Winner of the International AI Hackathon 2025 for connecting low-latency patient monitoring with role-aware clinical workflows and immediate Telegram escalation.
- Delivery
- Prototype, 2025
- Focus
- Backend, ML
- Scope
- 6 core technologies
Ownership and evidence
What I built
The product decisions and engineering work that shaped the final result.
Structured the FastAPI backend around independent inference, identity, persistence, and notification services.
Built a low-latency YOLOv8 inference path for continuously evaluating camera frames for patient anomalies.
Implemented authenticated doctor, nurse, and administrator workflows with Supabase-backed role enforcement.
Connected detections to Telegram alerts with defensive failure handling and operational logging.
Product evidence
See it in action
Interface views and demonstrations from the working product.
Visual assets are not published for this project. The verified delivery and engineering scope are documented above.