Quick update on [BioStat Quest](https://biostatquest.com) — case-based biostatistics trainer with FSRS-6 spaced repetition. I shipped the full Pro pillar set this week and wanted to put it in front of people who'd actually use it.
What's free vs Pro
Free tier covers the foundational arc of every branch (~3 cases per branch), the FSRS-6 scheduler, layer-1 hints, AI tutor (5 questions/week), 2 timed exams per month, top-3 misconception ledger.
Pro adds: full catalog (50+ cases, 1,000+ questions), unlimited AI tutor, full hint ladder, unlimited exam mode with branch-level breakdown, the printable Statement of Competency, and the full per-tag misconception drill-down with practice routing.
Open beta caveat: payment processor activation is still in review, so signed-in users get all Pro features for free right now. Paid plans launch shortly — monthly $9 / yearly $60. Early-beta users get a launch discount.
---
What actually shipped this week
- AI tutor as a real chat. Visible message thread, scoped to one question at a time, refuses off-topic. Quick re-asks for "explain simpler", "give an example", "show the formula". Each turn is single-shot server-side (no multi-turn context = no jailbreak surface), but the client accumulates the thread for visibility. Free quota: 5/week. Pro: unlimited.
- Statement of Competency with public verification. Print-ready PDF (browser print dialog → save as PDF). Each statement has a unique Doc ID (FNV-1a hash of email + ISO date + per-tier counts). When you generate a statement, the client posts an idempotent record server-side. Anyone can verify it at `biostatquest.com/verify` — paste the Doc ID + issuer's email and the endpoint confirms authenticity (returns "Verified · issued by [initials] on [date], covering N branches"). Same response shape for "not found" and "wrong email", so existence isn't disclosed.
- 3-layer hint ladder. Layer 1 (orienting nudge) is free for everyone. Pro unlocks Layer 2 (names the structural component you're missing) and Layer 3 (partial walkthrough). The hint panel shows L1/L2/L3 dots so you can see the curriculum at a glance.
- Misconception ledger → practice queue. The ledger tracks the patterns in your wrong answers (the question's authored misconception tag matches the distractor you picked). New: each tag now surfaces the cases that contain questions targeting that misconception — click to jump straight into focused practice instead of just reading "you fell for X".
Happy to take feedback. Particularly interested in:
- Which Pro feature would matter most for your workflow?
- Anything in the verify flow that feels off?
- Methods/cases you'd like to see added to the bank.
