MedicSense AI
Link to open source: https://github.com/aakash4dev/medisence-ai
MedicSense AI is a comprehensive health assistant designed to help users take charge of their health through three core pillars:
Multimodal AI Diagnostics
-
Natural Language Symptom Check: Users describe their symptoms in simple, everyday language. By integrating Google Gemini 1.5 Pro, the system analyzes context, duration, and severity to deliver an immediate preliminary assessment.
-
Vision AI Analysis: Users can upload images of skin conditions or injuries. The AI evaluates the visual data to detect possible issues, assess severity, and recommend appropriate first-aid steps.
Smart Triage & Safety
The platform categorizes health concerns into clearly defined levels—Mild, Moderate, Serious, and Emergency. Based on this classification, it provides personalized guidance, from home-care suggestions for minor issues to specialist recommendations, and activates Emergency Protocols (such as formatting alerts to call 911) for critical cases.
Seamless Connectivity
-
Instant Booking: Users can schedule doctor appointments directly within the app.
-
WhatsApp Integration: Using Twilio, real-time WhatsApp notifications are sent to doctors with patient details immediately after booking, minimizing administrative delays.
⚙️ How We Built It (Tech Stack)
-
Frontend: HTML5, Vanilla CSS (Glassmorphism Design System), JavaScript (ES6+).
-
Backend: Python (Flask).
-
AI Engine: Google Gemini 1.5 Pro (via
google-generativeaiSDK) for both NLP and computer vision tasks. -
Notifications: Twilio API for WhatsApp messaging.
-
Data & Authentication: Firebase along with Local Storage / JSON.
✨ Key Features
-
🩺 AI Doctor in Your Pocket: Interprets medical context beyond simple keyword searches.
-
📷 Image Recognition: Lets users show their symptoms instead of only describing them.
-
🔔 Real-time Alerts: Doctors receive instant WhatsApp notifications.
This build was uploaded as a hackathon project




