# Salaah Reminder Team

- **Event:** [Built with Opus 4.6: a Claude Code hackathon](https://cerebralvalley.ai/e/claude-code-hackathon)
- **When:** Feb 10 at 12:00 PM – Feb 17 at 10:00 AM (EST)
- **Where:** Location TBA
- **Team:** [Abubakkar Rangara](https://cerebralvalley.ai/u/aar)
- **GitHub:** https://github.com/aarangara/salaah-reminder-app
- **Demo video:** https://youtube.com/shorts/W0tqwTNGiuw?feature=share
- **Gallery:** https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery/112

Our app allows user to see the prayer times of mosque they follow -  and VERY accurately. Users search their mosque or scan its timetable - which I beleive none of the apps support - with AI-powered OCR, then receive accurate notifications. Works fully offline.

Claude Haiku 4.5 vision extracts structured prayer times from any timetable format—handling diverse layouts, language variations, and mixed time formats. Extracts 5 daily prayers (start + jamaat), expands date ranges, and provides confidence scores. Auto-corrects time ambiguity.

Claude Opus 4.6 was used to build this complex project - Full-stack monorepo - expo/rn mobile app, Hono/PostgreSQL backend, React admin panel. Offline-first with SQLite, end-to-end type safety, local notifications, auto-sync.

Will work for any mosques & help 1.8B Muslims globally - you just need to snap a photo.

Search → select mosque → scan new timetable with camera → Claude extracts in seconds → save → instant notifications → offline mode.

---

Markdown version of https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery/112. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt