My1Health: Redesigning Digital Health Infrastructure for Medical Tourism
End-to-End Application
Enterprise Platform
Business to Business (B2B) – Referral Partner Network System (RPNS) – 9 weeks
Highlight Overview
My1Health connects patients travelling abroad for medical treatment with partner hospitals overseas — a market where 70% of potential patients were dropping out during signup and document collection (surfaced in an interview with My1Health founder and CEO, Ryan Marincowitz). The internal tool used to manage this process, the Referral Partner Network System, relied on heavy manual typing, forced every case through identical steps, and buried Medical Tourism Officers (MTOs) — staff who each manage up to 50 active patient cases at once — in dense, error-prone screens.
I led the redesign of the client intake and profile-creation flow, with one target: dramatically cut MTO manual workload while making the system fast enough to use under real operational pressure. The redesign centred on three key changes:
- Document collection moved to the very start of the process, rather than being an afterthought later in the flow.
- OCR (Optical Character Recognition) auto-extraction — technology that automatically reads text from images or scanned documents, like a passport or ID, and converts it into editable, searchable digital data — so officers never have to retype identity details manually.
- A location engine, the Regional Proximity Network Selection (RPNS), that automatically ranks nearby partner hospitals by travel distance, safe post-surgery flight windows, and visa processing time — replacing work MTOs had previously been doing by hand across separate tools.
Impact:
Validated through unmoderated usability testing with 12 real users:
- 100% found and used core navigation without issue
- 95% completed the new registration flow without getting stuck
- 88% completed the new document-upload step successfully
Formally reviewed by My1Health’s CEO, Ryan Marincowitz, who:
- Called RPNS “the most ambitious conceptual design thinking”
- Put six core features directly onto the company’s product roadmap — including RPNS, predictive stall alerts, and the OCR-driven Secure Vault Bridge (projected to cut manual entry by 78%)
The Challenge
The medical tourism market is growing, but a major bottleneck was costing My1Health patients before they ever reached a hospital: 70% of potential patients dropped out during initial signup and document collection (source: CEO interview).
The legacy system compounded the problem — manual data entry, one rigid workflow for every case regardless of urgency, and crowded, text-heavy screens were causing eye strain and slowdowns for officers already managing high caseloads, while giving them no fast way to identify the right partner hospital for a given patient.
My Role & Method
Role: Product Designer & UX/UI Specialist — Strategy, Interaction Design, Accessibility Audit, Prototyping & Testing
Tools: Figma, Gemini AI, Chat GPT, Lyssna
Before designing any screens, I identified everyone the system touches — not just the officers using it day-to-day, but the managers relying on its data and the overseas hospitals receiving the files it produces:
- Medical Tourism Officers (MTOs) — the day-to-day operators who’d actually live in this tool managing up to 50 concurrent cases, needing fast, scannable workflows.
- Country Managers — track booking volume and travel-corridor trends (e.g. UK → Turkey).
- Clients (patients) — the people the whole system exists to serve, even though they’re not the ones logging into RPNS.
- Partner Desks — the overseas hospital staff on the receiving end of whatever file the system produces who need clean, complete files to make fast care decisions.
I then ran a structured interview with founder/CEO Ryan Marincowitz to identify the sharpest operational friction points. This shaped the entire redesign priority list:
- Missing or outdated patient records
- The 70% document-collection drop-off
- Passport/record name mismatches causing delays
- Visa “showstopper” deposits – a large upfront deposit some destination countries require as part of the visa application — often to prove the traveller can financially support themselves, or as a guarantee tied to the medical treatment itself.
- Highly varied technical confidence among referral partners — many preferring WhatsApp over the internal system
The Solution
I restructured the intake process around four principles:
- Streamline (cut visual clutter, plain-language labels)
- Localise (auto-match patients to the best partner hospital using travel corridors and visa timelines)
- Automate (OCR document capture moved to the start of the process not the end)
- Customise (routine vs. urgent cases automatically follow different, appropriately prioritised workflows).
The redesigned flow
Intake & Identity → Flexible Document Capture → Location Match → Final Review — replaced a slow, one-size-fits-all process with one that adapts to case urgency and does the location/logistics reasoning automatically:
- Flexible document capture
Officers text a secure upload link to the patient’s phone, or upload files themselves. Either path triggers automatic OCR extraction with a real-time accuracy score — no manual retyping, no more name-mismatch errors. - Automatic hospital matching (RPNS)
Officers used to manually research travel distance, safe post-op flight windows, and visa processing times across separate sources. Now the system ranks best-fit hospitals the moment a patient’s home country and treatment type are entered.
Example: for a patient travelling from Bangladesh, the system surfaces Thailand and India as top matches — based on proximity, visa corridor, and flight-safety timing.
The Goal: guided choice, not just choice. A static directory became a small, ranked, context-aware shortlist. - Error-safe final review
The review screen is split into independently editable blocks — profile, hospital, logistics. One mistake no longer means redoing the whole form. - Built for real-world stress, not just the happy path
Testing showed stressed clients often delay submitting documents. Two features responded directly to that finding (not part of the original plan):- A 48-hour stall alert that flags and resurfaces stuck cases automatically
- An “Urgent Logistics” panel that only appears when a trip is inside its critical 48-hour window
- Accessibility built in, not bolted on
- Contrast fixes for low-vision/high-glare conditions
- 48px-minimum tap targets
- Jargon replaced with plain language, since many officers are non-native English speakers
(viewing larger image will open in new tab)
Validation
I ran the wireframes through unmoderated remote testing on Lyssna with 12 participants before moving to high-fidelity design:
- 100% found and used the core navigation without issue.
- 95% completed the new registration flow without getting stuck.
- 88% completed the new document-upload step successfully.
Testing didn’t just confirm the direction — it changed the design. The 48-hour stall alert and the Urgent Logistics panel both came directly out of what testers revealed about how they’d actually behave under pressure, not out of the original plan.
One tester’s comment captured exactly what the redesign was reaching for:
“It made the complex registration feel like a conversation. I didn’t feel like I was filling out a medical form. I felt like I was helping a person.”
Business Outcome
This project was presented and handed to My1Health’s founder and CEO, Ryan Marincowitz, who provided direct written feedback afterward. His response is clear evidence that my design work translated into real business decisions — six features went straight onto the company’s product roadmap:
- RPNS location-matching logic — Ryan called it “the most ambitious conceptual design thinking” noting the England → Turkey matching example “was exactly the kind of insight that lands with our team”.
- Predictive stall-state alerts (48-hour amber / 72-hour red) — described as “exactly the kind of operational tooling that turns a CRM into an operating system.
- OCR-driven Secure Vault Bridge — SMS-triggered document capture with a projected 78% reduction in manual entry, which Ryan flagged as “a specific testable hypothesis, which is rare.
- “Fetch My Client” functionality — for cases where a Country Manager has already assigned a client — which he noted “solves a real workflow gap we hadn’t articulated.
- Three-path personalisation (generalist / urgent / concierge, with the concierge path supporting patients travelling with family).
- Accessibility audit baseline — built on the WCAG contrast work below.
Need to elevate your digital project?
Let’s discuss how my design system, accessibility, and product strategy skills can help your team.