Lab values photographed, interpreted and turned into two reports.
Austrian health-tech company offering AI-supported health analysis.
- 2 reports
- from the same data, for patient and doctor
- 4 tiers
- from questionnaire to epigenetic testing
- GDPR
- role-based access to health data
- Computer vision
- Language models
- Patient app
Illustrative images created with AI. The projects and their results are real.
The challenge
The client wanted to turn a health check into a digital product: patients answer a medical questionnaire, add their lab results, and receive an analysis they understand, while doctors get the detail they need.
Lab results arrive as photos and PDFs from different labs, health data is sensitive, and no model may make medical decisions on its own.
What we built
A platform made of several applications: a guided questionnaire app, a patient portal, a cloud backend with an admin panel, and a small bridge service, plus the public website that sells the four service tiers.
OCR turns lab documents into structured values; a language-model pipeline writes a plain-language report for the patient and a clinical version for the doctor to review.
What it does
- Guided digital health questionnaire, managed by the client's team in the admin panel
- Lab results read automatically from photos and documents (OCR)
- Two reports from one analysis: plain language for the patient, clinical for the doctor
- Lab value history with trend charts
- Self-registration with email verification and role-based access
- Marketing website presenting the four service tiers
The result
Doctors review a structured clinical draft for every patient, and patients understand their results in plain language.
The client owns a platform that grows tier by tier, from the basic check to lab values, medical records and epigenetic testing.
How we built it
Laravel backend with a REST API and a Filament admin panel; Ionic and Angular for the questionnaire app and the patient portal; AWS storage, Redis, real-time updates, email and SMS.
The AI pipeline combines OCR with language models and leaves medical decisions to clinicians. The website runs on Next.js, built mobile-first for speed and accessibility.
Our role
Full-stack development of the backend, apps and website, design of the AI pipeline and technical GDPR advice, delivered in phases.
Technology
- AI agents
- RAG
- Mistral
- n8n
- Microsoft Presidio
- OCR
- Laravel
- Filament
- Ionic
- Angular
- AWS