BREAKING
Sample data — this page shows preview athlete data while the platform is in beta. Real athletes onboard at learn.html or via a Rep 1 combine.
System Demo

The 15 Seconds — Live

See the athlete identification and clip capture system in action.

Camera Demo Active
Rep 1 Exposure Camera Station

Enable your camera to see the athlete identification system in action. No data is stored.

Camera access stays on your device. Nothing is uploaded in this demo.

MW
Marcus Williams
WR · Class 2027 · Lincoln HS
✓ ID
Recording — 0:00 / 0:15
Demo mode — uses simulated detection sequence
01
Scan Zone Active

Camera station watches the drill zone continuously. Face recognition runs against enrolled athlete database.

02
Athlete Identified

Face matched to enrolled athlete. ArUco marker on bib serves as backup. System locks the 15-second rolling buffer.

03
Profile Updated

Clip uploaded to Azure Blob. Profile updated in real-time. Pinned scouts notified. Social export queued.

Video Pipeline
Camera Station
iPhone / laptop running OpenCV + face-api.js
Edge Processing
30-second rolling buffer · ArUco + face detection
Azure Blob Storage
SAS token upload · direct from edge device
Event Grid Trigger
Fires on blob created event
Azure Function
Transcode · tag · notify scouts
Athlete Profile
Clip live within 60 seconds of capture
Social Export Queue
Vertical crop for TikTok/Instagram (opt-in)
Identification Methods
🎯
Primary — Face Recognition
Pre-enrolled athletes. 3 selfies in app before event. AWS Rekognition / Azure Face API in production.
🏷️
Secondary — ArUco Marker
Printed bib for unenrolled athletes. Robust to motion blur and angle. OpenCV native detection.
🔢
Fallback — Jersey OCR
AWS Rekognition text detection reads jersey numbers. Event-scoped deduplication handles same-number collisions.
View Sample Profile → Scout Search →