Nyrocare · Shri

The problem

$300B lost to medication non‑adherence every year

Nearly half of US adults with hypertension don't have their blood pressure under control. Not because they don't care, but because they're overwhelmed, unsupported, and left alone between clinic visits. Existing digital health tools fall into two camps: too clinical (medication reminders with no context) or too generic (wellness apps that ignore the complexity of chronic disease management).

The gap

No one connects the dots

Patients track medications in one app, food in another, and vitals in a third. No tool combines all three with clinical intelligence that actually understands drug‑nutrient interactions or condition‑specific thresholds.

The opportunity

Clinical depth + empathetic design

What if a chronic disease companion could combine evidence‑based clinical guidance, personalized tracking, and an AI health companion that understands your specific conditions?

Competitive analysis

7 platforms analyzed, one clear gap

I evaluated the major digital health platforms for chronic disease management. Each has strengths, but none combines medication intelligence, condition‑aware nutrition, and conversational AI in a single experience.

💊

Medisafe

Strong reminders, no nutrition or AI companion

🍎

MyFitnessPal

Generic nutrition, no clinical awareness

🩺

Livongo / Teladoc

B2B payer model, limited self‑serve

🤖

Ada / Babylon

Symptom triage, no ongoing chronic management

Nyrocare's moat

The real competitive advantage isn't code. It's clinical domain expertise: deep knowledge of treatment protocols, drug‑nutrient interactions, and condition‑specific thresholds, combined with empathetic design that makes people feel supported rather than surveilled.

The solution

Three pillars in one app

Nora

AI Health Companion

Voice‑enabled AI that answers health questions grounded in peer‑reviewed evidence. Understands your conditions, medications, and history. Not a chatbot. A clinical copilot.

Clinical engine

Smart medication tracking

Integrated pharmaceutical database with drug interaction alerts, adherence patterns, and condition‑specific medication intelligence from trusted clinical sources.

Nutrition AI

Condition‑aware food logging

Vision ML food recognition with condition‑specific nutritional analysis. Knows that potassium matters differently for someone on ACE inhibitors vs metformin.

My role

Full product lifecycle, one seat

This wasn't a team project where I owned one slice. I led every phase: problem research, competitive analysis, clinical hypothesis development, customer persona creation, research design (qualitative interviews + quantitative survey targeting 100+ respondents), value proposition canvas, MVP scope, technical architecture, app development, brand identity, and go‑to‑market strategy.

From clinical hypothesis to working app. Research, product, engineering, brand, and strategy in one seat.

Skills applied

User researchProduct strategy Technical architectureApp development Competitive analysisBrand design Clinical research

Tech stack

React NativeExpo Node.jsVector DB Voice AIVision ML GitHub

Journey

10 months, 4 phases

May to Oct 2024

Research and discovery

Problem research, competitive analysis across 7+ platforms, clinical literature review, persona creation, 50+ user interviews, 12‑section living research document.

Nov 2024 to Jan 2025

Brand and architecture

Brand setup (domain, logo, LinkedIn, YouTube), developer recruitment, technical architecture decisions, infrastructure planning, database schema.

Feb to Mar 2025

Build and ship

Voice AI companion, clinical knowledge engine, intelligent food logging, pharmaceutical data integration, 19 integrated services, full app audit passed with 0 errors.

Current

Polish and pilot

Knowledge retrieval system, onboarding flow, dashboard polish, preparing for pilot launch with target users.

User research

What 50+ interviews taught me

The research phase wasn't a checkbox. It fundamentally changed the product direction. Three insights reshaped every design decision:

01

Trust requires clinical grounding. Users rejected generic AI health advice instantly. They trusted the app only when answers cited specific clinical sources they could verify.

02

Tone is a feature. The difference between "You missed your medication" and "Ready for your evening dose?" is the difference between surveillance and support. Retention depends on this.

03

Integration, not features. Users didn't want more apps. They wanted one place that connects medications, food, and vitals with intelligence about how they interact for their specific conditions.

So what

What this demonstrates

Dimension	What I did
Problem discovery	50+ user interviews, competitive analysis of 7 platforms, clinical literature review
Product strategy	Persona creation, value proposition canvas, MVP scoping, go‑to‑market planning
Technical execution	React Native app, 19 integrated services, voice AI, vision ML, vector DB for clinical knowledge
Clinical rigor	Evidence‑based design grounded in treatment protocols, drug interaction databases, condition‑specific thresholds
Brand and launch	Full brand identity, domain, social presence, audit passed with 0 errors, pilot in preparation

F‑1 compliance note

Nyrocare was developed as a graduate venture project through the WashU Skandalaris Center for Entrepreneurial Studies, fully compliant with F‑1 visa curricular practical training requirements.

Tech Stack

Built with

Mobile‑first stack with Claude Code as the primary implementer. Every feature shipped from a written spec, locked decisions, and on‑device verification.

React Native TypeScript (strict) Expo Supabase MMKV Claude Code Figma Jest Git