Skip to content

v21.0: Admissions Portfolio & External Validation Plan

Status: PARKED behind v22 Gate 0 / Gate 1, but now fully triaged as an admissions evidence backlog
Priority: Strategic, but subordinate to v22.0 non-duplicate-value proof
Time Horizon: 12 weeks once v22 allows external execution
Last Updated: 2026-04-04

Why This Plan Exists

This document is the admissions-focused roadmap for turning CareConnect from a technically strong pilot into something that reads as serious leadership, community impact, and professional execution on OMSAS ABS / CanMEDS without exaggeration.

The core rule is unchanged:

  1. Do not over-package weak evidence.
  2. Do not prioritize style/polish over real implementation, governance, or adoption.
  3. Do the work that creates defensible external artifacts first.

Current State

  • CareConnect is live at https://careconnect.ing and has unusually strong technical/governance scaffolding for a student-led project.
  • The main unresolved question is no longer "can it ship?" but "does it create non-duplicate real-world value relative to 211?"
  • The highest-value gap is external execution: named pilot partners, real workflow use, measurable outcomes, provider-confirmed records, and external validation.
  • The repo already contains significant pilot, governance, privacy, accessibility, and observability infrastructure.
  • The admissions story will improve most by producing real pilot evidence, not by adding more product breadth or cosmetic polish.

Triage Rules

Use these rules whenever two backlog items compete:

  1. Work that directly closes v22 blockers outranks everything else.
  2. Work that produces real-world outcome evidence outranks packaging or presentation work.
  3. Work that strengthens safety, governance, verification, or pilot reliability outranks broadening the feature set.
  4. Public claims must lag evidence, not lead it.
  5. Optional prestige work (applications, incorporation framing, publications) only makes sense after core pilot evidence exists.

Tracking Conventions

Use these status labels when turning this plan into active execution tracking:

  • pending: accepted backlog item, not yet started
  • active: currently in execution
  • blocked: cannot progress because a named dependency is still unresolved
  • complete: artifact exists and is strong enough to cite honestly
  • dropped: intentionally removed because the evidence case became weak or redundant

Default assumption today:

  1. Tier 0 items are pending unless they have already been completed elsewhere in v22 tracking.
  2. Tier 1+ items are effectively blocked by the current v22 gate state unless they can be executed in a tightly bounded way without violating current roadmap rules.

Tier 0 Status Snapshot (2026-04-01)

  • A1: blocked pending user-owned Gate 0 evidence (UA-1, UA-3).
  • A3: complete via extended pilot metric-source schema, recompute-to-snapshot flow, docs alignment, and passing focused tests.
  • A11: complete via public claim/placeholder hardening on active repo-facing surfaces.
  • A22: complete via focused pilot/privacy route coverage and updated schema/metric tests.
  • A6: active in bounded mode because scoped readiness audit/export tooling now exists, but no real verification cycle artifact has been executed yet.
  • A16: active in bounded mode because scoped readiness reporting now exists, but pilot-scope data fixes are still pending.

Implication:

  1. The remaining near-term Tier 0 execution is A1 plus any bounded A6 / A16 follow-through that materially improves pilot readiness.
  2. A3, A11, and A22 should now be treated as baseline capability, not open backlog.

Priority Tiers

Tier 0: Do Now Only If It Directly Supports v22

These are the only admissions-relevant items that should move before v22 Gate 0 / Gate 1 evidence exists:

  1. A1 Close v22 Gate 0 with named pilot partners and clause-level legal/API evidence.
  2. A6 Run a bounded verification/freshness cycle if it directly supports partner trust or pilot readiness.
  3. A16 Close pilot-scope data quality gaps if they affect real pilot reliability.

Completed Tier 0 groundwork now already in place:

  1. A3 pilot metric computability
  2. A11 claim / placeholder hardening
  3. A22 focused pilot/privacy test hardening

Tier 1: First Work Immediately After Gate 0 Exit

These are the first execution items once external pilot activity is allowed:

  1. A2 Run a small real-world pilot in one defined referral workflow.
  2. A14 Run a formal crisis-safety validation protocol with provider review.
  3. A8 Conduct structured professional usability / fitness-for-referral sessions.
  4. A23 Create partner/referrer collateral to support adoption.
  5. A13 Formalize the CareConnect-to-211 boundary and handoff workflow.
  6. A7 Start targeted L3 provider-confirmation outreach.
  7. A9 Recruit the first advisory reviewers and hold an initial meeting.
  8. A27 Add a quick-exit / safe-leave flow for violence-sensitive browsing contexts.
  9. A28 Expose referral-critical logistics already present in the data model, such as service area, virtual delivery, languages, and transit notes.
  10. A29 Build shortlist / compare / multi-print referral workflows for referrers and partner staff.

Tier 2: First 2-6 Weeks of Live Pilot Evidence

These are the highest-value "this is real" proof artifacts:

  1. A4 Publish the first baseline-to-pilot scorecard and decision memo.
  2. A5 Secure at least one factual support / endorsement letter.
  3. A12 Publish a verification/freshness/transparency page.
  4. A21 Run a public status page and archive monthly operational snapshots.
  5. A20 Complete an incident-response drill and short post-mortem.
  6. A15 Perform manual accessibility testing with assistive-tech users.
  7. A17 Complete human-reviewed French/plain-language support for sensitive services.
  8. A18 Run the evidence-based identity/equity tagging pass.

Tier 3: Evidence Packaging After Initial Pilot Proof Exists

These are valuable, but they should follow actual pilot evidence:

  1. A10 Obtain an external privacy / AI safety review.
  2. A24 Make leadership/collaboration visible with named roles and real governance ownership.
  3. A26 Present to community or professional audiences.
  4. A25 Publish a case study, poster, or preprint.

Tier 4: Conditional / Context-Dependent Work

These are only worth doing if the surrounding context is real and timely:

  1. A19 Obtain an institutional QI / REB / privacy classification.
  2. A25 Formal dissemination, if the pilot data is strong enough to support it.
  3. A26 Formal external presentations, if there is an audience with actual relevance.

12-Week Artifact Stack

If the goal is for CareConnect to look like a serious, audited, adopted project within 12 weeks, these are the target artifacts to create in order:

  1. Dated Gate 0 closure evidence bundle.
  2. Pilot protocol with named scope, partner, workflow, and safeguards.
  3. Computable pilot metric stack with passing tests.
  4. Verification cycle log and pilot-scope data-quality delta report.
  5. Crisis validation report and professional usability summary.
  6. First pilot scorecard and decision memo.
  7. First L3 confirmations and at least one support letter.
  8. Advisory review minutes and external privacy/AI review memo.
  9. Transparency page, status page, and monthly ops snapshot.
  10. Referrer brief, demo asset, and either a presentation deck or case study.

Comprehensive Improvement Register

ID Improvement Current Priority Earliest Window Why It Matters Most Key Artifact(s) Dependencies / Notes
A1 Close v22 Gate 0 with named pilot partners and clause-level legal/API evidence Tier 0 Now Converts this from a strong repo into a governed external initiative Redlined terms, partner list, outreach evidence bundle, updated gate checklist Directly blocked on UA-1 and UA-3
A2 Run a small real-world pilot inside one actual referral workflow Tier 1 After Gate 0 exit Creates the strongest possible shift from prototype to real implementation Pilot protocol, workflow map, partner confirmation, pilot log Requires permission to run external pilot activity
A3 Finish pilot instrumentation so M2/M4/M5/M6/M7 are computable Tier 0 Now Outcome evidence is stronger than feature breadth Schema/docs updates, passing tests, non-null pilot scorecard fields Must preserve privacy guardrails
A4 Publish the first baseline-to-pilot scorecard and decision memo Tier 2 After initial pilot data This is the highest-value scholarship artifact in the backlog Scorecard, dated memo, threshold review Depends on A2 and A3
A5 Secure one or more factual support / endorsement letters Tier 2 After external review or early pilot use Independent validation matters more than self-description Letter on letterhead or official email Best requested after concrete use/review
A6 Execute a recurring verification cycle and publish the results Tier 0 Now if it supports pilot readiness Operationalizes the strongest governance claim: accuracy over breadth Verification log, correction summary, updated freshness stats Safe to start early if bounded to pilot relevance
A7 Achieve the first real L3 provider-confirmed services Tier 1 After Gate 0 exit Proves the governance model is functioning in the real world Confirmation emails, updated verification counts, screenshots Outreach should follow claim-flow testing and partner readiness
A8 Conduct structured professional usability / fitness-for-referral sessions Tier 1 After Gate 0 exit Demonstrates evidence-based iteration with credible users Session protocol, notes, findings summary, resulting fixes Needs access to front-line professionals
A9 Recruit advisory reviewers and hold the first meeting with minutes Tier 1 After Gate 0 exit Shows real oversight, not just drafted governance Named members, meeting agenda, minutes, action tracker Start with bounded asks, then formalize
A10 Obtain an external privacy / AI safety review Tier 3 After initial pilot proof External ethical scrutiny is much stronger than internal audit alone Review memo/email, response log, updated public summary Better after pilot scope is concrete
A11 Audit public claims, placeholders, and stale messaging Tier 0 Now Unsupported claims erode credibility faster than missing features Claim-evidence matrix, updated pages, before/after screenshots Includes placeholders in acknowledgments and partner-facing pages
A12 Publish a transparency page for verification, freshness, corrections, and limits Tier 2 After first executed ops cycle Shows humility, accountability, and safety culture Public transparency page, dated snapshots Stronger after A6 and early pilot evidence
A13 Formalize CareConnect’s boundary with 211 and build the handoff workflow Tier 1 After Gate 0 exit Demonstrates system judgment and reduces overreach risk Boundary statement, UX copy, handoff flow, partner feedback Should align with v22 positioning work
A14 Run a formal crisis-safety validation protocol with provider review Tier 1 After Gate 0 exit Addresses the highest-risk user pathway directly Crisis scenario sheet, screenshots, issue log, reviewer sign-off High ethical priority
A15 Do manual accessibility testing with real assistive-tech users Tier 2 During early pilot Real accessibility evidence is stronger than automated checks alone Manual test report, fixed issues, updated accessibility statement Needs tester access
A16 Run a pilot-scope data completeness sprint Tier 0 Now if pilot-critical Missing coordinates/hours/email/freshness weakens real-world reliability Before/after completeness report, updated audit outputs Scope to pilot services first, not repo-wide polish
A17 Complete human-reviewed French/plain-language support for sensitive services Tier 2 Early pilot Strengthens the equity claim in a concrete way Completed fields, reviewer note, bilingual/plain-language gap report Focus on crisis/shelter/DV/high-risk services first
A18 Run an evidence-based identity/equity tagging pass Tier 2 Early pilot Shows ethical rigor and avoids performative tagging Evidence table, updated records, audit summary No vibe-based tagging
A19 Obtain an institutional QI / REB / privacy classification Tier 4 When human-subject boundary becomes real Demonstrates ethical process awareness Determination email or approval letter Needs clarification on institution / oversight body
A20 Run an incident-response drill and publish a short post-mortem Tier 2 Early pilot Shows operational maturity and safety rehearsal Drill report, post-mortem, runbook updates Can be simulated without waiting for a real failure
A21 Add a public status page and monthly ops snapshots Tier 2 Early pilot Signals accountable stewardship and continuous monitoring Status page, archived monthly snapshots, uptime screenshots Best once there is a pilot audience to serve
A22 Harden automated tests around crisis, pilot, privacy, and claim-critical flows Tier 0 Now Improves core reliability rather than surface polish Targeted tests, CI evidence, focused coverage gains Keep scope on highest-risk flows
A23 Build partner/referrer collateral that makes adoption easy Tier 1 After Gate 0 exit Helps external professionals evaluate and actually use the tool One-page brief, demo asset, onboarding pack, LOI/MOU template Should reflect true current state only
A24 Make leadership and collaboration visible, not implied Tier 3 After first real collaborators exist Leadership is stronger when it is documented and shared Updated roles page, contributor/governance ownership matrix, meeting cadence Do not fabricate team scale
A25 Publish a case study, poster, or preprint Tier 3 / 4 After real pilot data exists Converts implementation into scholarship Poster, abstract, preprint, slide deck Needs clarification on venue and authorship
A26 Present CareConnect to community or professional audiences and archive the feedback Tier 3 / 4 After initial proof artifacts exist Demonstrates communication and stakeholder engagement Invitation, slides, attendance note, feedback summary Needs clarification on audience access
A27 Add a quick-exit / safe-leave flow for violence-sensitive browsing Tier 1 After Gate 0 exit Reduces harm for users who may be searching in unsafe environments Quick-exit UX, copy review, crisis-safety validation notes Align with A14; keep behavior discreet and keyboard-accessible
A28 Expose referral-critical logistics on public service pages Tier 1 After Gate 0 exit Referrers care about access constraints as much as descriptions Updated detail UI, professional-usability findings, before/after screenshots Align with A8; use existing vetted fields only
A29 Build shortlist / compare / multi-print workflows for referrers Tier 1 After Gate 0 exit Makes partner adoption easier in real referral settings Shortlist UX, printable packet flow, partner/referrer feedback Align with A23; only build what pilot users actually need

Immediate Sequencing Recommendation

If the project wants the highest admissions value without getting ahead of itself, the execution order should be:

  1. A1, then bounded A6 / A16.
  2. A2, A14, A8, A23, A13, A27, A28, A29, A7, A9.
  3. A4, A5, A12, A21, A20, A15, A17, A18.
  4. A10, A24, A26, A25.
  5. A19 only if the external oversight question becomes real.

This sequence deliberately avoids polishing-first work and keeps the strongest admissions gains tied to governance closure, actual use, and measurable outcomes.

If Working On This Today

Do not pick randomly from the register. Use this order:

  1. A1 if external evidence needed for Gate 0 is available to process.
  2. A6 and A16 only where they improve pilot trust/readiness, not as generic cleanup.
  3. Otherwise, hold the admissions backlog until Gate 0 permits Tier 1 execution.

Success Markers

This backlog is working if it produces:

  • one real pilot scope with named ownership,
  • one outcome scorecard with non-null values,
  • one verification log,
  • first L3 confirmations,
  • one external support letter,
  • one advisory/external review artifact,
  • one public transparency surface,
  • one status / ops evidence surface,
  • and one dissemination artifact grounded in real results.