CareCompass AI

Inspiration

The idea for CareCompass AI came from a simple but common problem: managing personal health can be confusing and overwhelming. Many people forget to take their medications, struggle to understand prescriptions, or keep medical documents scattered across different places. I wanted to build something that makes healthcare information easier to understand and manage, especially for people who don't have immediate access to medical professionals or who simply want to stay organized.

Rather than creating another AI chatbot, I wanted to build a practical health companion that could assist users in their day-to-day healthcare journey.

What it does

CareCompass AI is an AI-powered personal health companion that helps users:

• Scan and organize prescriptions using OCR.
• Explain medical terms and prescriptions in simple language.
• Set medication reminders and refill alerts.
• Track symptoms and monitor health trends over time.
• Store medical documents securely in one place.
• Generate concise health summaries to prepare for doctor visits. The goal is not to replace healthcare professionals but to help users better understand and manage their health information.

How we built it

The application was built with a modern full-stack architecture:

• Frontend: Next.js, TypeScript, Tailwind CSS, and shadcn/ui for a responsive and clean user experience.
• Backend: FastAPI (Python) to power the API and AI workflows.
• Database & Storage: Supabase for authentication, PostgreSQL database, and secure document storage.
• AI: OpenAI/Gemini for natural language understanding and medical explanations.
• OCR: Document scanning to extract prescription details automatically.
• *Charts & Analytics: Visual dashboards to display medication adherence and symptom history. The system processes uploaded prescriptions, extracts important information, organizes it into structured records, and uses AI to provide easy-to-understand explanations while helping users stay on track with reminders and health insights.

Challenges we ran into

One of the biggest challenges was designing AI interactions that are genuinely helpful without giving medical advice. I had to carefully craft prompts so the assistant focuses on education, organization, and guidance while encouraging users to consult healthcare professionals for diagnosis or treatment.

Another challenge was handling prescription documents, since different formats, handwriting styles, and image quality can affect OCR accuracy. Creating a workflow that transforms extracted text into meaningful, structured information required several iterations.

Balancing a feature-rich experience with an intuitive interface was also important. I wanted users to access reminders, health records, and AI assistance without feeling overwhelmed.

Accomplishments that we're proud of

What we learned

Building CareCompass AI taught me a lot about combining AI with real-world workflows instead of using AI as a standalone feature. I gained experience integrating OCR with large language models, designing secure document management systems, and building user experiences that prioritize simplicity and trust.

More importantly, I learned that successful AI applications aren't just about intelligent models—they're about solving real problems in ways that are practical, accessible, and responsible.

What's next for CareCompass AI

This project is only the beginning. Future improvements include:

• Voice conversations with the AI assistant.
• Integration with wearable devices such as Apple Health and Google Fit.
• Family caregiver accounts for managing loved ones' medications.
• Multi-language support.
• Offline reminders.
• Personalized health insights based on long-term trends.
• Secure sharing of health summaries with healthcare providers. My vision is to make CareCompass AI a trusted digital companion that empowers people to take better control of their health while complementing—not replacing—professional medical care.

Built With

• next.js
• react