MediFlow

Inspiration

Managing multiple medicines can be confusing-different food instructions, timing gaps, and medical terms leave patients unsure what to take and when. We built MediFlow to turn this confusion into one simple, clear schedule.

What it does

MediFlow converts multiple medication instructions into a personalized daily timeline, detects potential timing conflicts, and explains the schedule in simple language.

How we built it

We combined a React frontend, backend APIs, OCR for prescription labels, and AI-based instruction parsing with drug-safety data to create the flow: Capture → Understand → Resolve → Schedule.

Challenges we ran into

The biggest challenge was handling multiple timing constraints without losing important instructions. OCR accuracy on real-world labels was another challenge, so we added manual entry as a fallback.

Accomplishments that we're proud of

MediFlow goes beyond reminders—it understands instructions and organizes them into a practical schedule, making complex medication routines easier to follow.

What we learned

We learned that healthcare AI requires accuracy, transparency, and safe failure handling. Even a simple medication schedule involves many real-world constraints.

What's next for MediFlow

We plan to add pharmacist/caregiver support, verified clinical databases, refill integration, and clinical validation to make MediFlow safer and more reliable.

Built With

• css
• express.js
• git
• html5
• if
• javascript
• json
• language
• llm
• natural
• node.js
• ocr
• openai-api-(or-anthropic-api-/-claude
• react
• tailwind
• tesseract
• that's
• use)
• what
• you