AI-based web app for disease prediction

Inspiration

In Morocco, genetic and lifestyle-related diseases often go undetected until late stages. Medical consultations can be costly, screening infrastructure is limited, and many patients lack reliable information about their own risk. our app was born from the idea that an accessible web app—powered by AI and grounded in local data—could help people understand, monitor, and act on their health risks long before symptoms appear.

What it does

Our app lets users securely enter basic health metrics (age, weight, habits), family history, and simple lab results. The app then:

Predicts disease risk (e.g., diabetes, hypertension, hereditary anemia) and displays a clear risk gauge.
Explains key drivers behind the prediction in everyday language.
Suggests preventive actions—diet tips, exercise goals, and when to consult a professional.
Stores results privately so users can track changes over time. ## How we built it
Data & Model
- Curated anonymized clinical data and open-source health datasets.
- Trained a classification model and layered it with an explainability module.
Architecture
- A lightweight API handles prediction requests, validation, and caching.
- The front end is responsive, mobile-first, and uses simple charts for clarity.
Security & Privacy
- All data is encrypted in transit and at rest; no personal identifiers are stored. ## Challenges we ran into
Limited local datasets required extensive cleaning and augmentation.
Balancing accuracy with explainability so users trust and understand results.
Data privacy concerns demanded strong encryption and transparent consent flows.
User engagement—turning technical risk scores into actionable, friendly guidance. ## Accomplishments that we're proud of
Delivered a fully functional prototype in under eight weeks.
Achieved risk-prediction accuracy on par with published benchmarks. ## What we learned
Data quality beats data quantity; small, clean, relevant datasets outperform large noisy ones.
Health UX matters—visuals and language must be friendly, not frightening.
Privacy is non-negotiable; clear policies build user trust.
Cross-disciplinary collaboration (medical, design, and AI) accelerates innovation. ## What's next for AI-based web app for disease prediction
Expand the dataset with regional hospital records (with proper consent).
Add voice-input and Darija language support for broader accessibility.
Integrate with wearable devices for continuous monitoring.
Pursue medical-device compliance and partner with public-health programs to scale impact.

Built With

css**-|-python/django-backend-and-js-+-html/css-frontend-are-described-as-core-stacks-:contentreference[oaicite:0]{index=0}:contentreference[oaicite:1]{index=1}-|-|-**backend-framework**-|-**django**-(mvt-architecture
css3
django
easy-prototyping)-|-the-report-notes-django-orm-managing-data;-sqlite-is-the-default-embedded-db-during-development-:contentreference[oaicite:6]{index=6}:contentreference[oaicite:7]{index=7}-|-|-**ide-/-code-editor**-|-**visual-studio-code**-with-python
form-handling
github
gitlens
html
html**
javascript
javascript**
numpy
orm
panda
prettier-extensions-|-chosen-for-all-coding-tasks-:contentreference[oaicite:8]{index=8}:contentreference[oaicite:9]{index=9}-|-|-**version-control**-|-**git**-(cli)-&-**github**-(remote-repository
python
responsive-interface-:contentreference[oaicite:4]{index=4}:contentreference[oaicite:5]{index=5}-|-|-**database**-|-**sqlite**-(default-django-orm-store
security)-|-detailed-in-chapitre-3-and-chosen-as-main-framework-:contentreference[oaicite:2]{index=2}:contentreference[oaicite:3]{index=3}-|-|-**frontend-styling**-|-**bootstrap**-for-responsive-ui-components-|-used-to-build-a-clean
sqlite
|-layer-/-purpose-|-technologies-&-tools-|-evidence-in-report-|-|-|-|-|-|-**languages**-|-**python**