HealthCare+

Healthcare+ is an AI-powered health companion that unites disease prediction along with fitness, nutrition, mental wellness, and medical care for smarter, simpler, and preventive living.

Comment

Healthcare+

Links

What We Were Building

We built Healthcare+, a multimodal, agentic AI-powered ecosystem that brings together every aspect of healthcare i.e physical, mental, and preventive into one intelligent platform. Our goal was to solve India’s fragmented healthcare system by creating a unified solution that empowers people to take charge of their well-being.

Core Features

  • Disease Prediction: Machine learning models predict potential health risks early, enabling preventive action.
  • Smart Nutrition Tracking: CNN-based computer vision estimates calories from food images.
  • Personalized Fitness Planning: AI crafts adaptive workout plans tailored to user goals and habits.
  • Period Tracker: Helps women monitor cycles, predict fertile windows, and manage hormonal health.
  • Find Clinics Nearby: Location intelligence helps users discover nearby hospitals and clinics.
  • Mental Wellness: PHQ-9 analytics for mood tracking and depression screening.
  • Medicine Management: OCR simplifies prescription tracking and medication reminders.
  • Multilingual Chat Interface: Offers AI healthcare support in multiple languages for inclusivity.

Built with Next.js, Supabase, LangChain, and multiple AI models, Healthcare+ transitions healthcare from reactive treatment to proactive wellness, making it more connected, intelligent, and human-centered.

How It Evolved

Week 1: Foundation and Core Setup

We began with the Next.js frontend and Tailwind CSS for responsive design. Supabase handled authentication and database management. The landing page and structure were built for scalability. OCR-based medicine tracking was our first working AI feature helping users manage prescriptions easily.

Week 2: Intelligence and Backend Integration

We focused on the backend using Express.js with proper CORS setup and environment configurations. Integrated the Gemini API for AI-based disease prediction and built a CNN food recognition model for calorie estimation giving Healthcare+ the intelligence to understand users’ inputs.

Week 3: Personalized Experience and Expansion

Introduced AI-driven workout recommendations, diet tracking, and period cycle monitoring. Created a detailed fitness schema considering age, gender, weight, and goals for truly personalized insights.

Week 4: Agentic AI, Accessibility, and Final Polish

Integrated LangChain for reasoning and multi-step task handling. Expanded multilingual support for rural users and optimized offline capabilities for low-connectivity areas. Ended with a UI/UX revamp based on real user feedback.

Key Technical Challenges Solved

  1. Secure environment and API management with dotenv.
  2. Reliable CORS and Axios communication between client and server.
  3. Improved OCR accuracy for prescription parsing.
  4. Optimized database schema for large, interrelated health data.
  5. AI pipeline orchestration using LangChain for agentic reasoning.
  6. Real-time CNN food estimation with faster inference speeds.
  7. Multilingual and offline support for inclusive healthcare.

Why It Matters

Addressing Real Problems

India’s healthcare is fragmented users rely on multiple apps for fitness, diet, and medical care. Healthcare+ brings everything under one roof, creating a holistic experience.

Rural Accessibility

With multilingual and offline capabilities, we’re bridging the gap between urban and rural healthcare — making quality health guidance available even without strong internet access.

Proactive vs Reactive

Instead of waiting for illness, Healthcare+ predicts risks early and provides preventive insights using continuous monitoring and AI-driven reasoning.

Mental Health Focus

PHQ-9 analytics help users track mood patterns privately, reducing stigma around mental health and encouraging early intervention.

Scalable Impact

The modular design can integrate with wearables, telemedicine, and national health systems — allowing impact at both individual and community levels.

The Vision

Healthcare+ represents a paradigm shift from fragmented tools to a unified, adaptive healthcare ecosystem. Through agentic AI reasoning and multimodal understanding, it learns, adapts, and scales to reach over 1.4 billion people, ensuring healthcare that’s personalized, preventive, and inclusive.

What Inspired Us

The inspiration came from witnessing how disconnected healthcare apps fail to provide holistic care. We wanted to build something that feels like a personal health companion smart enough to predict issues, simple enough for anyone to use, and empathetic enough to care about the whole person, not just symptoms.

What We Learned

We learned how different AI domains computer vision, natural language processing, and reasoning models can come together to solve real-world problems. We also realized that good healthcare design isn’t just technical it’s emotional, requiring empathy, clarity, and trust. A surprising takeaway: debugging healthcare AI can teach patience better than meditation.

Chaos Diary Highlights

  • 3 AM debugging: dotenv refused to load forgot to restart the server. The real bug was sleep deprivation.
  • OCR thought “Paracetamol” was “Parachute oil.” For a day, Healthcare+ became a skincare brand.
  • CORS errors everywhere until we realized we were testing on the wrong port.
  • CNN estimated pizza at 2000 calories instead of 2800. Our AI clearly loves comfort food.
  • LangChain agents froze mid-demo perhaps they achieved temporary sentience.
  • Supabase refused to sync the Wi-Fi had quietly switched to hotspot mode. Lesson: always check the network first.

Funny Bugs

  • Disease predictor’s universal solution: “Drink water.” Technically not wrong.
  • Fitness planner suggested 100 push-ups for beginners; tough love, AI style.
  • Period tracker notified male users, gender filter vanished mid-update.
  • Chatbot’s comforting message: “Everything will be fine, bro.” Slightly too relatable.
  • Calorie tracker confused a cup of tea with ramen, maybe it was just hungry.

Built With

Share this project:

Updates