Health Score App
Development Plan
Phase 1 & Phase 2 Comprehensive Roadmap
Privileged & Confidential
01 — Executive Summary
Vyten will design and build a Health Score web app that aggregates 6 health input categories into a single 0–100 composite score, embedded via WebView into native mobile apps, starting with the Arete Wellness pilot brand and rolling out to Mantality Health, BioDesign, and the Agentis parent app. Phase 1 uses admin-panel-inputted data with an integration-ready architecture; Phase 2 connects real data sources sequentially.
| Milestone | Date | Status | Notes |
|---|---|---|---|
| Project Kickoff | Feb 4, 2026 | Complete | Discovery session with Jimmy, Rachel, Ashlyn |
| Updated Wireframes Delivered | Feb 11, 2026 | Complete | Ultrahuman style, Arete branding, actual markers |
| Full Development Plan Delivered | Feb 12, 2026 | Complete | Comprehensive roadmap, Gantt, dependency tracker, multi-brand strategy |
| Board Meeting Demo | Mar 10, 2026 | Sprint 2 Target | Working web dashboard with Arete branding + mock data |
| Arete MVP in Production | May 8, 2026 | Phase 1 Target | UAT-tested, App Store submitted |
| Multi-Brand Rollout Begins | May–Jun 2026 | Phase 2 | Mantality, BioDesign themes applied |
| Phase 2 Integrations Complete | Jan 2027 | Phase 2 | All live data sources connected |
Architecture Decision: The app uses a hexagonal (Ports & Adapters) architecture with mock data adapters in Phase 1. This means the full user experience gets built now — and when real integrations arrive, we swap adapters without touching core application code. Development velocity is fully decoupled from external dependency resolution.
02 — Dependencies & Risk
MindWise will NOT be used per Feb 4 decision. Evaluating MindWise, Vibrant, AdvancedMD.
Target: Apr 1, 2026
Charlie / Ben Smarr finalizing composite scoring model. Was due Jan 31, reported "On Track."
Target: Mar 15, 2026
All data flows through Agentis-owned Databricks. Environment must be provisioned by Agentis IT.
Target: Apr 15, 2026
Partnership with Ultrahuman for BloodVision lab panel "in progress." API access, data ownership TBD.
Target: May 1, 2026
Agentis must create Apple Developer + Google Play Developer accounts for app deployment.
Target: Feb 14, 2026
OneDrive folder shared Feb 10. Reviewing markers, data types, visualization preferences, trending reqs.
Received: Feb 10, 2026
Mitigation Strategy: Every dependency has a mock/fallback adapter in the architecture. Development proceeds at full velocity regardless of dependency status. When a dependency is resolved, the real adapter is swapped in — zero impact on timelines for work already completed. Each month a dependency is delayed past its target date, the corresponding Phase 2 integration month shifts accordingly.
Probability: High | Impact: High
Mitigation: Build EMR-agnostic data abstraction layer using FHIR R4 as internal standard. Research all 3 EMR APIs now. Present API comparison to accelerate decision.
Probability: Medium | Impact: High
Mitigation: Build pluggable scoring engine. Use placeholder weighted-average model. Swap real algorithm when finalized.
Probability: Medium | Impact: Medium
Mitigation: Design lab panel for multiple data sources. Consider Terra API as middleware. Support CSV import as fallback.
Probability: Medium | Impact: Medium
Mitigation: Build against local PostgreSQL with matching schema. Migration to Databricks = configuration change, not rewrite.
03 — Timeline
04 — Sprint Detail
05 — Phase 2 Roadmap
| Month | Integration | Dependency | Doc Quality | Est. Effort | Status |
|---|---|---|---|---|---|
| Month 4 May 2026 |
Questionnaire Digitization WPHI + COSEHC-17 + PHQ-9 as patient-facing forms |
Algorithm specs from Charlie | Self-contained | 2–3 weeks | Ready |
| Month 5 Jun 2026 |
InBody / Grip Strength Body composition + grip strength via LookinBody API |
InBody account credentials | Good | 3–4 weeks | Awaiting Access |
| Month 6 Jul 2026 |
Ultrahuman / Junction Wearable data + BloodVision lab panel (74 markers) |
API access + partnership finalized | Moderate | 4–6 weeks | In Negotiation |
| Month 7 Aug 2026 |
Databricks Pipeline Push/pull health data from Agentis data platform |
Databricks environment provisioned | Good (when ready) | 3–4 weeks | Awaiting Provisioning |
| Month 8 Sep 2026 |
EMR Integration Connect selected EMR for patient chart data |
EMR vendor selected + API access | Varies by vendor | 4–10 weeks | Decision Pending |
| Month 9 Oct 2026 |
Cognivue Integration Cognitive function test results (5 domains) |
Cognivue Portal access + export format | CSV only | 3–4 weeks | Awaiting Docs |
| Month 10–11 Nov–Dec 2026 |
Multi-Brand Rollout Mantality + BioDesign production deployment |
Brand assets + location configs | Self-contained | 2 weeks each | Phase 2 |
| Month 12 Jan 2027 |
Analytics + Feature Enhancements Event tracking, dashboards, cohort recommendations |
Analytics stack decision | Self-contained | Ongoing | Phase 2 |
Re-sequencing Rule (per SOW): The order above is the recommended sequence based on integration readiness, not the original SOW order. Integrations are completed sequentially unless mutually agreed otherwise. If a higher-priority dependency is resolved before a lower-priority one, we pull it forward. Each delayed dependency pushes its integration month — not the entire project.
| Factor | AdvancedMD | MindWise Health | Vibrant Practice |
|---|---|---|---|
| API Documentation | Full developer portal | None public | None public |
| FHIR R4 Support | Yes (read-only, free) | Unknown | Unknown |
| Sandbox Access | Yes | No | No |
| Integration Pathways | 4 (Connect, FHIR, ODBC, HIE) | Custom (vendor engagement) | None documented |
| Clinical Fit | General practice | Behavioral health focus | Functional / longevity medicine |
| Integration Timeline | 4–6 weeks | 4–8+ weeks | 3–6+ months |
| Scorecard Total (from deck) | 410 | 423 | 337 |
Recommendation: Regardless of which EMR is selected, the Health Score app is being built with FHIR R4 as the internal data standard. This means AdvancedMD integration would be straightforward, MindWise would require direct vendor API engagement, and Vibrant would initially use CSV/manual import. Surfacing this analysis to Jimmy and Rachel now will help accelerate the decision.
06 — Multi-Brand Strategy
| Deliverable | Effort | Owner | Notes |
|---|---|---|---|
| Brand theme CSS file | 2–4 hours | Vyten | Colors, typography, spacing mapped to design tokens |
| Brand config TypeScript | 1–2 hours | Vyten | Display name, logo URL, feature flags, legal text |
| Logo & visual assets | Provided by client | Agentis / Brand | SVG logo, favicon, app icons |
| Feature flag configuration | 1–2 hours | Vyten | Enable/disable brand-specific modules |
| Visual regression tests | 2–4 hours | Vyten | Automated screenshot comparison across key screens |
| App Store listing + submission | 1–2 days | Vyten | Screenshots, description, metadata per brand |
| Total per brand | 1–2 weeks |
| # | Category | Data Source | Phase 1 | Phase 2 |
|---|---|---|---|---|
| 1 | Whole Person Health Questionnaire | 9 questions (social/emotional wellbeing) | Admin input → patient-facing form | Digital assessment in-app |
| 2 | COSEHC-17 Assessment | CV risk questions + lab markers | Admin input | Digital assessment + lab integration |
| 3 | Grip Strength | InBody device (highest of 3 attempts) | Admin input (manual) | InBody WebAPI integration |
| 4 | Comprehensive Lab Panel | 74+ markers via Ultrahuman / LabCorp | Admin input (manual) | Ultrahuman Junction API |
| 5 | Cognitive Function | Cognivue — 5 domains (memory, attention, processing, executive, visuospatial) | Admin input | Cognivue CSV import → API (if available) |
| 6 | Body Composition | InBody device (body fat %, muscle mass, BMR, BMI, body water) | Admin input | InBody WebAPI integration |
07 — Action Items & Decisions
| Action | Owner | Due | Status |
|---|---|---|---|
| Share OneDrive folder with markers spreadsheet, API docs | Rachel Scott | Feb 10 | Complete |
| Whitelist Vyten domains for OneDrive access | Rachel Scott (IT) | Feb 10 | Complete |
| Create Apple Developer Account | Rachel / Ashlyn | Feb 14 | Due This Week |
| Create Google Play Developer Account | Rachel / Ashlyn | Feb 14 | Due This Week |
| Finalize Health Score algorithm with MAB | Jimmy / Rachel | End of Feb | In Progress |
| Finalize Ultrahuman lab contract + connect API contact | Jimmy / Rachel | Feb 14 | In Progress |
| Schedule call with Charlie (algorithm) + Chris (Databricks) | Rachel / Michael | Feb 14 | Pending |
| Deliver updated wireframes (Ultrahuman style, Arete branding) | Spencer / Vyten | Feb 11 | Complete |
| Provide wireframe feedback (48-hr SLA) | Agentis Team | Feb 13 | Due This Week |
Tuesdays at 1:30 PM CT — 30-minute standup led by Michael Turner. Written recap within 24 hours. Dependency tracker reviewed, blockers escalated, sprint progress reported. This is the single forum for all project communication. Agentis POC: Rachel Scott. Vyten POC: Michael Turner.
Development is fully underway with a working prototype targeted for the March 10 board meeting. The Phase 1 architecture is designed so that no dependency delay can stop development progress — every feature is built against mock data, and real integrations are swapped in when ready. The early May production target for the Arete Wellness pilot is on track. Phase 2 integrations follow at one per month, with sequencing that adapts to which dependencies are resolved first.
The single highest-impact decision remaining: Selecting the go-forward EMR vendor. This decision affects the integration timeline more than any other factor.