Symptom Checker
Tracker
Blood tracker
Perscription analyser
Health tracker
Hospital Locator
Landing Page

🩺 CareNav.ai – AI-Powered Health Companion

CareNav.ai is a unified, multi-module healthcare ecosystem designed to bridge the gap between symptom onset and clinical intervention. By combining generative AI, machine learning, and OCR, it acts as a preventive digital assistant to reduce medical misinformation and improve health literacy.

🌍 Inspiration

Healthcare accessibility is often hindered by delayed guidance, misinterpreted symptoms, and fragmented follow-ups. We observed five core pain points:

The "Google Panic": Users searching symptoms online often face extreme, inaccurate self-diagnoses.
Prescription Barriers: Difficulty reading handwritten notes or understanding dosage instructions.
Adherence Issues: High rates of forgotten medications and inconsistent vital tracking.
Data Fragmentation: No centralized place to monitor sugar, BP, and pulse trends.
Emergency Navigation: Lack of quick, specialty-based hospital locating in critical moments.

🧠 What We Built

CareNav.ai moves beyond a single-feature model to provide a complete healthcare workflow:

Module	Functionality
AI Symptom Triage	Conversational health interaction powered by Gemini.
Smart Prescription Analyzer	OCR-based interpretation of medicine names and dosages.
Health Tracker Dashboard	Longitudinal monitoring for Sugar, BP, and Pulse.
Medicine Reminders	Automated alerts to improve medication adherence.
Hospital Locator	Specialty-based filtering (optimized for Tamil Nadu regions).

⚙️ How We Built It

1️⃣ AI Symptom Analysis

Integrated the Gemini API for sophisticated reasoning. The pipeline follows: User Input → AI Reasoning → Risk Classification (Low/Med/High) → Actionable Guidance.

2️⃣ Disease Prediction Model

We trained a supervised ML model on structured symptom datasets. Reliability was validated using the Accuracy formula:

$$Accuracy = \frac{TP + TN}{TP + TN + FP + FN}$$

3️⃣ Prescription & Vital Management

OCR Integration: Image upload functionality extracts text from prescriptions, which the AI then "translates" into plain language.
Data Visualization: A health dashboard transforms raw vital logs into visual trends, aiding in continuous health monitoring.

4️⃣ Full Stack Architecture

Frontend: Interactive UI with multilingual support (English, Hindi, Tamil).
Backend: High-performance API layer using FastAPI.
Database: MongoDB for flexible and secure health data storage.
Security: Implemented encrypted communication and strict input validation.

🚧 Challenges & Learnings

Building in the healthcare space requires balancing innovation with safety:

Safety First: Ensuring AI responses remain non-diagnostic and provide necessary medical disclaimers.
Data Integrity: Managing OCR inaccuracies from low-quality prescription images.
User Experience: Designing a UI that reduces medical anxiety rather than fueling it.
Key Insight: Healthcare AI is about responsible guidance, not just raw prediction accuracy.