# Detextit

- **Event:** [Gemini 3 SuperHack](https://cerebralvalley.ai/e/gemini-3-superhack)
- **When:** Sat, Jan 31 at 9:00 AM – 10:00 PM (PST)
- **Where:** 365 Toni Stone Xing, San Francisco, CA
- **Team:** [Sarath Shekkizhar](https://cerebralvalley.ai/u/shekkizh)
- **GitHub:** https://github.com/detextit/cv3-gemini-hack
- **Demo video:** https://www.youtube.com/watch?v=iAKZlBtOP5w
- **Gallery:** https://cerebralvalley.ai/e/gemini-3-superhack/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/gemini-3-superhack/hackathon/gallery/12

detextit is an AI-powered basketball gameplay analysis tool that uses Google Gemini Flash's agentic capabilities to perform tactical breakdowns of uploaded video or image clips. It renders real-time SVG overlays, attack lines, defense rotations, zone regions, and player movement paths, directly on top of gameplay footage, translating raw AI insights into visual strategy diagrams. Built with React, Vite, and the Google GenAI SDK, the app runs an autonomous multi-iteration reasoning loop where Gemini continuously generates and refines tactical annotations (progressively rendered) until a thorough analysis is complete. The result is a simple app showcasing this capability. Latency to first render varies from ~4s on smaller screens  and ~15s on larger screen.

Website: https://cv3-gemini-hack.vercel.app/

---

Markdown version of https://cerebralvalley.ai/e/gemini-3-superhack/hackathon/gallery/12. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt