Veritas — Lumina

One-line summary: Lumina helps women understand lab results, decode clinical reports, and prepare for conversations with doctors.

HER Hackathon Track

• Track: Health
• Team Name: Veritas
• Team Members:
  • Olumide Michelle Oluwanifemi — Developer
  • Adedunye Imisioluwa Praise — Web Developer
  • Ewelike Virginia Oluchi — Researcher

Problem Statement

Women often struggle to interpret their lab reports because medical results are written in technical language and lack personalized context. This leads to confusion, anxiety, and missed opportunities to act on health signals.

Evidence from validation:

• Early feedback showed users felt lost reading numerical lab values without plain-language explanations.
• Users wanted a trusted way to convert report findings into doctor questions and actionable summaries.
• People preferred a mobile-friendly, accessible experience with both visual and audio guidance.

Solution Overview

Lumina transforms lab result PDFs into easy-to-understand health explanations, trend analysis, and personalized doctor advocacy checklists. It also provides audio narration so users can absorb insights hands-free.

Core features:

1. PDF report upload with OCR extraction and structured marker parsing
2. Medical result decoding plus trend chart analysis across reports
3. Personalized doctor question checklist and audio summary playback

Demo

• Live Demo: https://lumina-frontend-one.vercel.app
• Demo Video: https://drive.google.com/file/d/1hO-z2ALSOTIRAyJzzA61QpJNGQaZEJPN/view?usp=drivesdk

Screenshots

Add screenshots or GIFs of your product here.

Screen	Description
	Upload and report interpretation screen
	Trend chart and advocacy checklist

How It Works

1. The user uploads a lab report PDF or image.
2. The backend extracts markers with Google Gemini and analyzes them with Groq.
3. The app shows simplified explanations, trends, and audio narration.

Validation & Research

Who we spoke to / researched:

• Health-conscious women exploring lab report clarity
• Peer mentors and product advisors in women's wellness
• Desk research on menstrual and reproductive health result interpretation

Key findings:

Finding	Evidence	Product decision
Lab results feel overwhelming	User feedback and interview notes	Add plain-language summaries for each marker
People want action steps	Observations from validation talks	Build a doctor question checklist feature
Audio helps accessibility	Early testers asked for read-aloud support	Add TTS fallback and browser audio playback

Tech Stack

• Frontend: React, Vite, TypeScript, Tailwind CSS
• Backend: Node.js, Express, TypeScript
• Database: PostgreSQL with Prisma ORM
• AI / API Tools: Google Gemini, Groq SDK, Cloudinary, YarnGPT TTS + browser speechSynthesis fallback
• Deployment: Vercel (frontend), Render or equivalent backend hosting

Architecture

[User Browser]
  ↓
[Frontend: React + Vite]
  ↓
[Backend: Express + Prisma]
  ↓
[PostgreSQL] / [Google Gemini] / [Groq] / [Cloudinary]

Installation / Setup

Prerequisites

• Node.js 18+
• PostgreSQL
• API keys for Google Gemini, Groq, and Cloudinary

Clone the repository

git clone [repository-url]
cd Lumina

Install dependencies

npm install
cd backend && npm install && cd ..
cd frontend && npm install && cd ..

Set up environment variables

Create .env files in backend/ and frontend/ if required.

Example backend variables:

DATABASE_URL=
CLOUDINARY_URL=
GOOGLE_API_KEY=
GROQ_API_KEY=
FRONTEND_URL=https://lumina-frontend-one.vercel.app

Example frontend variables:

VITE_API_URL=https://<your-backend-domain>/api

Run locally

# Backend and frontend must run separately in this repo structure
cd backend && npm run dev
cd ../frontend && npm run dev

Open http://localhost:5173 in your browser.

Usage

1. Upload a lab report PDF or image.
2. View the decoded health markers and trend charts.
3. Listen to the audio summary and copy doctor questions.

Project Structure

.
├── backend/
│   ├── prisma/
│   ├── src/
│   │   ├── index.ts
│   │   ├── lib/
│   │   ├── middleware/
│   │   ├── routes/
│   │   └── services/
│   └── package.json
├── frontend/
│   ├── public/
│   ├── src/
│   │   ├── components/
│   │   ├── hooks/
│   │   ├── lib/
│   │   ├── pages/
│   │   └── types/
│   └── package.json
└── README.md

Challenges We Faced

• Extracting medical markers accurately from PDFs and report images
• Making result explanations clear without oversimplifying health context
• Ensuring audio narration works reliably with backend TTS and browser fallback

What We Would Improve Next

• Add more report formats and lab provider templates
• Expand support for personalized cycle and hormone guidance
• Add user accounts and secure saved history across devices

Business / Sustainability Model

• Users/customers: women seeking better lab report clarity and health confidence
• Revenue model: subscription or premium access to deeper explanation features
• Key partners: clinics, telehealth providers, wellness coaches
• Main costs: hosting, AI API usage, data privacy and support

Team Contributions

Name	Role	Contribution
Olumide Michelle Oluwanifemi	Developer	Frontend and backend implementation
Adedunye Imisioluwa Praise	Web Developer	UI development and integration
Ewelike Virginia Oluchi	Researcher	User research and health content validation

Acknowledgements

• NITHUB, University of Lagos
• HER Hackathon mentors, facilitators, judges, and volunteers
• Users and stakeholders who helped us validate the problem

License

For hackathon/demo purposes only.