
SoGoal AI
2024 — PresentSoccer data analytics AI ecosystem & real-time broadcast graphics pipeline: 22-player optical tracking, sub-second pitch calibration, and Unreal Engine 5 virtual advertising.
- Computer Vision
- Unreal Engine 5
- Player Tracking
- Homography AI
- Virtual Advertising
Co-Founder and Chief Technology Officer (CTO) at SoGoal AI, directing the research, development, and system architecture of our end-to-end soccer computer vision and broadcast graphics ecosystem.
Built around two proprietary core AI engines:
• Full-Pitch Optical Tracking AI: High-precision computer vision pipeline (SkillCorner-grade) that detects, tracks, and extracts 2D pitch coordinates, velocity vectors, and physical metrics for all 22 players and the ball across broadcast video feeds.
• Automated Pitch Calibration & Homography AI: Instantaneous camera pose estimation and lens distortion correction that maps multi-angle broadcast camera feeds onto calibrated 3D world coordinates in milliseconds.
Unreal Engine 5 Broadcast Integration: Connected our tracking and calibration pipelines directly to Unreal Engine 5 via low-latency data streams to render dynamic 3D virtual pitch advertisements, live tactical passing lanes, player ID heatmaps, and telemetry overlays that perfectly lock to the pitch with full camera motion parallax.
Web Analytics Platform: Designed and deployed sogoalai.com, providing football clubs, coaches, and sports broadcasters with comprehensive interactive performance dashboards, match tactical maps, and automated video clip generation.
