Sprint 1 Deliverable · Last updated 2026-05-11

Lesson & Remediation
Pathways — the plan.

Phase 2 turns the existing OpenLiteracy MVP into an AI-driven adaptive learning engine that adjusts in real time to each student — and each group. This is the master plan. Sign it off, and we build.

Read the plan Browse the 116 open questions
3
sprints × 2 weeks
116
open questions
10
decision themes
6/21
target ship
01 — Vision

Build the Ferrari, not the Fiat.

Phase 1 shipped a working tool. It runs in classrooms today. But it was built fast, and the seams show: lesson logic is hardcoded, remediation is repetition rather than adaptation, group sessions are treated like 1:1 sessions, and the data the AI could use to make better decisions often doesn't get captured.

Phase 2 is six weeks to fix that — without breaking what works. We spend the first sprint on planning. We map every assumption, surface every ambiguity, and lock the architecture before a line of new code ships. That's the Ferrari. Sprints 2 and 3 build it.

This site is Sprint 1's deliverable. It's the plan, the open questions, the matrix of what AI does vs what humans do, and the schedule. Read it. Push back where it's wrong. Sign off when it's right.

OpenLiteracy
Sarah Scott Frank · Rebecca Patterson · Greta Phillips Kendall
Subject-matter, vision, sign-off.
Futureproof Studio + Ceridwen
Eric Schneider · Ryan Shuken · Beth Kukla
Build, integration, project management.
02 — Themes

10 decisions, then we build.

Every one of the 116 questions falls into one of these ten themes. Each theme has a single owner and a target sprint. Resolve them in order, and the engine designs itself.

03 — Timeline

The 6-week schedule.

Sprint 1 May 11 – 22
Foundation
50 questions resolved. Architecture locked.
  • • Settle all source-of-truth conflicts (N)
  • • Lock the 4-part lesson template (C.1, C.2)
  • • Design group sparse-data architecture (P)
  • • Initial assessment + transitions spec (A, B)
  • • AI model, voice, word-pool choices (L.1–L.5)
  • • Schema gap inventory (M)
Sprint 2 May 25 – Jun 7
Engine Room
40 questions resolved. Core logic built.
  • • Per-lesson authorship + content pipeline (C.3–C.11)
  • • Remediation logic (D)
  • • Progress monitoring + acceleration (E)
  • • AI generation flows (L.6–L.9)
  • • Group operations runbook (H)
  • • Time variants 15 / 25 / 30 min (J)
Sprint 3 Jun 8 – 21
Surface & Edges
26 questions resolved. UI, polish, integration.
  • • Educator-facing Learning Plan view (I)
  • • Foundations Mastery + Nonfiction Navigators (F)
  • • Roots / Level Zero for EL students (G)
  • • Brand + visual polish (K)
  • • Schema hardening (M.6 security)
  • • Integration tests + handoff
⚠️ Critical milestones
2026-05-22
Sprint 1 exit
Master plan signed off by Sarah. Engine design ready.
2026-05-29
Rebecca lock
All Rebecca-owned questions must be answered. She's in London Jun–Jul.
2026-06-21
Ship target
Phase 2 ready for OL classroom deployment.
04 — AI ↔ Human

Who does what.

Every capability in this system is tagged with one of four labels. Drawing the line clearly is part of the master plan — it's how we avoid building the wrong kind of AI feature.

05 — Open Questions

All 116 in one place.

The complete registry, filterable. Each question shows who owns it, what sprint it lands in, and how it's categorized.

Sprint
Owner
Bucket
Showing 116 of 116 questions.
ID Section Question Bucket Owner Sprint
06 — For You

Per-person reading lists.

The team has organized questions by who's best positioned to answer them. Find your name and you'll know where to focus.

07 — What's next

Next moves, by role.

Sarah
Adjudicate the 9 Source-of-Truth Conflicts.
Beth will send you a clean email with the 7 Section-N questions plus 2 strategy items this week. Reply with answers — Flyn will route them into the plan.
Rebecca
76 questions, target lock 2026-05-29.
You're the source of truth for curriculum logic — most pathway, transition, and remediation decisions live with you. Beth will batch these into 3 themed emails so you're not staring at 76 in one go.
Greta
14 UI / brand questions + mockup deliverable.
The Learning Plan view (Section I) is greenfield — your mockups are blocking. Target: initial pass by Sprint 2 week 1 (2026-05-29).
Eric & Ryan
17 schema + model questions; surface and resolve.
Sections L (AI generation) and M (data schema) are your domain. Don't wait on OL — resolve what you can internally, surface only what genuinely needs OL input.
Beth
Draft, send, chase. Approve Flyn's drafts.
Flyn drafts emails per stakeholder; you approve before sending. Daily 8am Telegram standup. Mid-sprint check-in with Sarah on Monday.
A note on this site. It will update as questions resolve. The canonical sources are in the GitHub repo — docs/02-open-questions/00_master-question-registry_Beth.md (registry), docs/02-open-questions/01_sprint-plan_Beth.md (schedule), docs/02-open-questions/why-we-ask.md (this overview), prior-code-analysis/ (MVP audit). When something's wrong here, fix the source; this page regenerates from it.