What attendance data powers
1. Grading readiness
Member profiles show classes logged at the current belt, so "are they ready?" starts from evidence. The judgement is always the professor's; the count is just honest.
2. Retention alarms
Sliding attendance is the earliest sign someone is drifting. The morning briefing flags at-risk members, and the AI can send a gentle nudge before the drift becomes a cancellation.
3. Instructor pay
Per-class pay rates count taught classes from the teaching log, which is built from real rosters and check-ins.
4. Timetable decisions
Analytics turns check-ins into class popularity and average fill, so you grow the slots that work and fix the ones that don't.
5. Credits and trials
Class packs and class-based trials decrement on check-in. No spreadsheet reconciliation; the tap is the ledger.
6. Settling disputes
"I was definitely there" and "I never trained that week" both end the same way: a timestamped record in the member's history and the audit log.