EpiNova

EpiNova: An intelligent platform that predicts fever outbreaks and assists real-time diagnosis empowering clinicians and communities to detect, diagnose, and prevent illness early.

Comment

Inspiration

Every monsoon season, hospitals across India overflow with fever patients — dengue, malaria, typhoid, viral flu — all sharing similar symptoms. We saw how delayed diagnosis and late outbreak detection cost lives and strain the healthcare system. We wanted to build a solution that predicts fever outbreaks before they happen and helps clinicians diagnose fevers faster and smarter.

What it does

EpiNova is an AI-driven fever intelligence platform that combines predictive outbreak modeling with AI-assisted fever diagnostics. It forecasts potential fever outbreaks across regions and provides clinicians with an intelligent triage tool to differentiate between types of fever based on symptoms, environmental data, and patient history — enabling early action and data-driven healthcare decisions.

How we built it

We integrated clinical data, environmental factors (temperature, rainfall, humidity), and real-time hospital reports into a central database. Using machine learning (XGBoost, Prophet), we trained models to detect outbreak patterns and predict risk zones. The backend was built with FastAPI and PostgreSQL, and the frontend dashboard with React and Tailwind CSS. Our AI triage engine uses natural language symptom inputs to suggest likely causes and next steps.

Challenges we ran into

  1. Data inconsistency across states and hospitals made model training difficult.

  2. Building an AI that could distinguish overlapping fever symptoms (like dengue vs. malaria) required fine-tuned datasets.

  3. Integrating real-time environmental APIs while maintaining model performance was tricky.

  4. Balancing accuracy with explainability to ensure clinicians could trust the AI’s suggestions.

Accomplishments that we're proud of

  1. Built a working AI model that accurately predicts outbreak trends with >80% accuracy in pilot regions.

  2. Developed an interactive dashboard for clinicians and public health officials.

  3. Created a smart fever triage assistant that reduces diagnostic uncertainty.

  4. Recognized by early healthcare partners for real-world impact potential in outbreak prevention.

What we learned

We learned that AI in healthcare isn’t just about prediction — it’s about trust and usability. Collaborating with medical professionals early ensures better model validation. We also learned the importance of clean, contextual data — good data beats fancy algorithms every time.

What's next for EpiNova

We plan to:

  1. Integrate with government surveillance systems for live outbreak alerts.

  2. Add wearable data and remote patient monitoring for personalized fever tracking.

  3. Deploy pilots in multiple states to validate scalability.

  4. Expand EpiNova into a full digital disease intelligence network for India and beyond.

Built With

Share this project:

Updates