Swasth Sarthi
Link to open source: https://github.com/merajsiddique07/Swasth-Sarthi
Project Description — Swasth Sarthi: Health’s Guide with AI
Swasth Sarthi is an AI-powered holistic healthcare platform designed to empower individuals and families to take charge of their physical, emotional, and preventive health.
It serves as a digital health companion — combining personalized tracking, pregnancy care support, instant medical access, and live health analytics in one seamless interface.
💡 Key Features and Functionalities
-
BabyCare Assistant:
-
Upload pregnancy checkup reports (PDF format).
-
Uses AI + OCR (pdf.js, Tesseract.js) to extract data and generate personalized insights like baby’s growth stage, week of pregnancy, nutrition tips, and safe activities.
-
Interactive mood selector and symptom-based recommendations.
-
-
InstantCall (Emergency Doctor Connect):
-
Enables users to connect with a certified doctor within 60 seconds via live video call using WebRTC technology.
-
Designed for accidents and medical emergencies where immediate pre-treatment guidance can save lives.
-
-
Live Health Report Analysis:
-
AI analyzes chat records and report data to detect health trends, generate wellness scores, and provide visualized analytics using Chart.js.
-
Offers users data-driven insights into their physical and emotional well-being.
-
-
To-Do Health Tracker:
-
Daily task list for physical exercises, medication reminders, hydration tracking, and wellness habits.
-
Helps users build healthy routines through consistent engagement.
-
-
AI Chat Assistant:
-
Integrated medical chatbot powered by OpenAI’s GPT-based API, enabling natural, conversational health queries and guidance.
-
⚙️ Technologies Used
-
Frontend: HTML5, CSS3, JavaScript, Font Awesome
-
AI & NLP: OpenAI GPT-4o model for intelligent report analysis and conversation
-
Document Processing: pdf.js, Tesseract.js for extracting data from medical reports
-
Data Visualization: Chart.js for generating wellness graphs and analysis charts
-
Backend (recommended): Node.js or Django for APIs, authentication, and data management
-
Video Communication: WebRTC for real-time video calling between user and doctor
-
Deployment: Render / Vercel for hosting frontend and backend services
🧠 Data Sources
-
Pregnancy checkup reports (PDF format) uploaded by users
-
User-generated chat history and symptom data
-
AI-generated insights and recommendations (via OpenAI API)
📊 Findings & Learnings
-
AI can simplify healthcare interpretation: Automated PDF analysis reduces dependency on manual medical interpretation.
-
User experience matters in health apps: A clean, emotional, and responsive design builds trust and engagement.
-
Integrating AI + WebRTC + OCR technologies creates a powerful hybrid system for real-time preventive healthcare.
-
Challenges faced: API key restrictions (CORS), integrating live video systems, and managing sensitive data securely.
-
Key learning: Combining empathy-driven design with AI-driven functionality results in meaningful digital health experiences.
This build was uploaded as a hackathon project
