The member
Can I make this visit happen?
A single thread connecting provider, cost, childcare, and the ride.
MYJOURNI / FEWER BARRIERS TO CARE
Make the appointment
a plan you can keep.
One care need.
One confirmed plan.
Book the whole morning.
Not just the appointment.
A provider who accepts your plan. A cost you understand. Childcare for the visit. A way to get there. myJourni connects these separate decisions into one coordinated care journey, so a booking becomes a plan someone can actually follow.
Explore the app demo- WHAT I BUILT
- Product strategy, UX & full-stack delivery
- COLLABORATION
- Maternal-health and payer-operations expertise
- THE PRODUCT
- Member experience & care coordination
01 / Context
Finding a doctor is only the beginning.
The starting point was a maternal-health access challenge: people were being asked to coordinate a fragmented system at precisely the moment they needed support. A provider directory could answer where to go. It could not resolve everything that stood between choosing a visit and attending it.
I reframed the work around a complete care need. The requirements had to account for time, cost, family responsibilities, transportation, and continuity with a care team, not simply the next available slot.
THE PROBLEM / 1:39
What stands between a need and care?
A shared vehicle
Build transportation around the confirmed time and location.
Children who need care
Reserve a childcare window that includes the visit and travel.
Uncertainty about cost
Show a coverage-based estimate before committing to a slot.
A disconnected care history
Request consent to reconnect the existing care team, with a preference-based fallback.
02 / People & workflow
One plan. Different people making it possible.
The care team
What needs attention before the visit?
Questions, visit context, and follow-up carried into the care journey.
The navigator
Who needs help before a missed visit?
Unresolved support arrangements and scheduling exceptions.
The support partner
Where, when, and for how long?
Reservations tied to the appointment and its real travel window.
The plan team
Is coordination changing access?
Care-need progress and source-defined evaluation measures.
THE SHARED CARE-NEED LIFECYCLE
- 1Raised
- 2Matched
- 3Priced
- 4Scheduled
- 5Supported
- 6Attended
- 7Followed up
03 / My contribution
The model, the product, and the story around it.
I carried myJourni from product definition through the materials that made the product legible: translating the coordination problem into requirements, an explicit domain model, conversation logic, interfaces, implementation, and a commercial story people could understand.
The pivotal distinction is Care Need versus Appointment. An appointment is a calendar commitment. A care need can hold the estimate, the booking, childcare, transportation, and follow-up. That is the foundation of a connected plan.
I authored the problem and solution scripts, created and directed the cloned AI voice, produced the After Effects films, and created the animation, transitions, pacing, and visual treatment. The featured care-navigation app began in Figma and now exists as a React-based product experience. Clinical content and journey validation remained collaborative responsibilities with subject-matter experts.
co-owned
Problem discovery and the coordination reframe
Problem statement, care desert research, provider distance and assignment analysis
owned
Business and journey requirements
Action-step and prompt requirement tables covering scheduling, cost, transport, childcare, what-to-expect, check-in, and follow-up
owned
Domain model and data contract
Preference and profile schema mapped onto the payer's USCDI v3 data classes; prior-payer care team import via interoperability
owned
System behavior and conversation orchestration
Prompt-by-prompt specification with branch logic, guided-option sets, and confirmation states
owned
UX design, end to end
Chat thread, provider card, slot picker, childcare confirmation, itinerary card
owned
Frontend implementation
Working prototype used for the pilot walkthrough
owned
Backend orchestration and integrations
Sequencing of match, cost estimate, scheduling, childcare and rideshare booking; email confirmation path
collaborated
Clinical content and journey validation
Subject matter expert review of trimester content, testing expectations, and medication-related guidance
influenced
Entry points, community partnerships, go-to-market
Six-channel entry point plan across community partners, retail, providers, agents, plan service, and digital
co-owned
Pilot design and funding case
Two-phase validation plan with scoped budget and duration
owned
Product narrative and commercial pitch
Defined the problem and solution story, wrote both film scripts, and shaped the case-study narrative around the product's coordination model
owned
Film direction and motion production
Created the AI-voiced narration, produced the After Effects films, and authored the animation, transitions, pacing, and visual treatment
owned
Product interface design and implementation
Designed the featured care-navigation app in Figma and carried the experience into its React-based implementation
THE DOMAIN MODEL
Member / Preference Profile / Care Team / Care Need / Provider/Offering / Cost Estimate / Appointment / Support Arrangement / Journey Stage / Check-in
04 / Options considered
The choices that defined the product.
The question was not how many features to add. It was which choices would make one complete care journey possible, and which complexity belonged outside the first release.
05 / Design decisions
Reduce uncertainty before asking for commitment.
Continuity before another search
Offer the existing care team when consent and data allow. Keep alternatives available instead of forcing a single recommendation.
Cost before the calendar
Make the estimate visible before a member chooses a slot. The estimate is a decision aid, not a guarantee of final responsibility.
Confirm a plan, not a fragment
Connect childcare and transport to the visit duration and location. If a support fails, preserve the appointment and expose the unresolved need.
Make the plan portable
Keep confirmations in the conversation and email. Offer reminders at each commitment rather than relying on the member to reopen the app.
Check-ins with accountable handoffs
Carry questions and context forward. Clinical review and explicit escalation boundaries remain essential; the interface is not a substitute for care.
Consent is part of the experience
Ask before retrieving prior-payer data. Make fallback matching understandable when consent or records are unavailable.
06 / The experience
One thread, from a need to a coordinated day.
THE SOLUTION / 1:46
See the product come to life.
detect the need
A short set of guided options; pregnancy, then a single clarifying question
Behind the experienceOpens a Care Need against her Member record and current Journey Stage
match
Her prior care team doctor named, with role, location, address and phone
Behind the experienceResolves Care Team from prior-payer interoperability data with consent; falls back to Preference Profile matching where absent
price
The cost of this visit under her coverage, stated before any dates
Behind the experienceGenerates a Cost Estimate against her benefits for the offering matched
schedule
Soonest available dates as chips, then time slots inside the chosen date, then a written confirmation with visit duration
Behind the experienceBooks the Appointment, records duration, emails a copy
arrange the supports
A childcare window with its reasoning stated, then a rideshare vendor choice, then a three-stop itinerary card
Behind the experienceCreates Support Arrangements chained to the appointment's real duration and location; emails each confirmation
remind
An inline opt-in at each commitment, and emailed details she can find without the app
Behind the experienceSchedules notifications against every element of the plan, not just the appointment
follow up
A post-visit check-in, the next appointment her care schedule calls for, and a stage-aware daily check-in between visits
Behind the experienceAdvances the Care Need state, opens the next one, carries the Check-in log into the next visit
The incomplete states are part of the product.
empty
No prior payer data or no consent — matching falls back to the Preference Profile and states that it has done so
exception
Childcare unavailable in the window — the appointment holds; virtual visit, alternative slot, or alternative partner offered; care need marked scheduled-but-unsupported
degraded
No rideshare availability — the appointment holds and transport surfaces as an unresolved exception rather than a failed booking
permissions
Explicit consent requested before any interoperability pull; consent recorded against the Member
audit trail
Every confirmation exists in the thread and in her email; past and future visits available in an appointment tracker
language
Profile-level language setting governs both matching and content, English and Spanish
07 / Evidence & value
Measure the plan that becomes an attended visit.
The supplied case study reports a 14-week pilot with 184 members in Miami-Dade and a matched comparison cohort. Its central evaluation question is whether coordinating the supports changes attendance, not simply whether the app produces more bookings.
REPORTED PILOT RESULTS / SUPPLIED CASE STUDY
87%
Kept-appointment rate
Compared with a reported 61% matched baseline.
6m 30s
Median time to a complete plan
Compared with a self-reported four-day baseline.
547
Confirmed care plans
From 612 reported care needs; 374 included support.
Figures reproduced from the supplied case study, not independently verified here. Source documentation contains conflicting pilot-provenance notes; reconciliation is pending. Read the source case study
Test the coordination hypothesis.
The reported kept rate was 91% with a support arrangement and 74% without one. That comparison motivates further evaluation; it does not, by itself, establish that the product caused the difference.
Keep the gaps visible.
The source also reports check-in completion declining from 71% to 44% and unmet requests for Haitian Creole. Retention and language access belong in the next iteration, not outside the success story.
MODELED COMMERCIAL SCENARIO / NOT REVENUE
$720K annual recurring revenue
The supplied commercial model assumes 100,000 covered lives at $0.60 per member per month across 25 groups. This is a scale scenario, separate from pilot observations and not a claim of realized sales.
08 / What comes next
A first journey. A broader coordination model.
Maternal health establishes a bounded lifecycle and a local support network. Expansion should prove that the same care-need model can support another journey without turning each new sponsor into a separate product.
The next layer is operational: role-specific views, partner availability, consent revocation, payer-data separation, auditability, and visible exception queues. New geographies and language packs need real service coverage behind them.
Deepen the first journey
Improve unsupported-plan resolution, cost-estimate accuracy, and engagement between visits.
Make the operating model repeatable
Configure sponsor rules and local partner networks without fragmenting the member experience.
Extend language access
Validate content and partner support together, including the language gaps reported in the pilot.
Validate a second care journey
Test the domain model against a different lifecycle before claiming a general-purpose platform.
HUMAN NEED / CONNECTED PRODUCT
Where does your
experience break apart?
I connect research, product decisions, and implementation around the work people need to complete.