Case study · 05
MedIntel
MedIntel helps registered users submit symptoms, receive possible-condition insights, track prediction history, and export AI-generated health reports.
- Role
- Full-Stack Developer
- Scope
- Frontend · API development · Database integration · AI-service integration
- Platform
- Web
- Year
- 2025
Overview
A health web app for symptom analysis, prediction tracking, and downloadable AI-generated reports.
The Challenge
The Challenge
People need a simple way to record symptoms and review possible health insights over time. MedIntel provides an authenticated workspace for submitting symptoms, viewing prediction history, and exporting results, while clearly presenting the output as non-diagnostic.
Approach
Approach
The build focuses on a core logged-in journey: account creation, symptom submission, prediction results, history, and profile management. It does not attempt to provide treatment, clinician workflows, or a comprehensive medical record system; the interface instead directs users to consult a healthcare professional.
Design
Design
The product uses a calm green health-focused visual system with light and dark modes. The dashboard prioritizes a health snapshot, quick links, recent predictions, a trend chart, and health tips, while the symptom checker keeps input constrained to searchable symptom chips, duration, and severity before analysis.
Engineering
Engineering
The Next.js App Router application exposes route handlers for authentication, profile management, predictions, dashboard data, history, and password resets. User and prediction records are persisted with Mongoose in MongoDB; passwords use Argon2 and sign-in issues JWTs. The analysis route converts ten selected symptoms to binary features, sends them to a configured FastAPI prediction service, stores the returned condition and confidence, and the client generates PDF reports with jsPDF and AutoTable. Resend and React Email support welcome and password-reset emails.
Outcome
Outcome
Live at medintel-web-app.vercel.app with the complete core flow working end to end: account registration, symptom analysis, persisted history, dashboard summaries, and PDF export.