Medinova

An AI-powered medicine companion that identifies pills, manages reminders, and supports users with empathy, clarity, and care — like a digital caregiver. Closer to your loved ones.

Comment

🧠 Inspiration

The inspiration for Medinova came from a very personal place — thinking about how my grandmother, like many elderly people, often forgets to take her medication on time or struggles with understanding how and when to take them properly. I wanted to build something that wasn’t just a set of alarms, but a compassionate, intelligent companion that could support real-world medication routines, even when someone doesn’t have the prescription or packaging at hand. That led me to combine the power of AI, visual pill recognition, and voice guidance into a unified experience that feels intuitive, informative, and friendly for users of all ages — especially seniors.

💊 What it does

Medinova is an AI-powered personal health companion that helps users manage their medications with confidence and ease. At its core, it functions as:

  • A smart medication reminder system (with snooze and taken tracking)
  • An AI chat assistant that answers medical queries in natural language (e.g., “Can I take this with milk?”)
  • A visual pill scanner that lets you identify medicines just by scanning the pill (shape, color, imprint — no packaging needed)
  • A diet and routine guide tailored to the medicine you're on, including food to avoid and ideal intake timing
  • A refill tracker that reminds users when their medicine stock is running low
  • A caregiver-friendly mode that allows loved ones to monitor and assist from afar
  • A clean, accessible UI with grandma-friendly features like large fonts, voice interaction, and emergency alerts

It’s not just a reminder tool — it’s an empathetic digital companion for medication management.

🔧 How I built it

I built Medinova using Bolt, which allowed me to rapidly prototype the AI interactions, structure the backend logic for reminders, and visually lay out the user experience.

Key components I implemented:

  1. A mock medication database with 6 elderly-relevant medicines (e.g., Paracetamol, Metformin, Thyronorm) including data like dosage, timing, food restrictions, and side effects
  2. A camera-based pill scanner, mocked using pill traits (color, shape, imprint) and matched to the database
  3. An AI chat assistant (powered by GPT-like intelligence) to help users navigate side effects, interactions, and routine advice
  4. A voice reminder system and voice-input compatibility for users who prefer speaking over typing
  5. A thoughtful UI/UX, optimized for elderly users with larger icons, contrast themes, and one-tap navigation

All of this was structured to work within Bolt’s ecosystem, using lightweight data handling, visual flow logic, and AI integration tools.

🧱 Challenges I ran into

Some of the key challenges I faced:

  • Balancing technical depth with elderly accessibility — building something intelligent without overwhelming the user
  • Designing the pill scanner logic in a realistic MVP scope (visual traits without real ML model training)
  • Creating an AI assistant that gives meaningful, medically-safe advice while staying simple and understandable
  • Keeping the UI focused — I had to resist the temptation to keep adding features and instead prioritize core functionality

🏆 Accomplishments that I'm proud of

  • Designing an app that combines empathy and AI — something both futuristic and truly useful for real-world users like my grandmother
  • Building a holistic medication experience that goes beyond static reminders to offer education, structure, and care
  • Making the app feel friendly and empowering, not clinical or overwhelming
  • Designing for accessibility by default — voice, visuals, emergency access, and simplicity

📚 What I learned

  • How powerful AI can be when contextualized properly in a user journey
  • That accessibility and elder care need to be top-of-mind when designing health tech
  • How to balance ambition with MVP reality, especially when building for people who might not be tech-savvy
  • The importance of empathy as a UX superpower — designing not just for tasks, but for comfort, understanding, and reassurance

🚀 What’s next for Medinova

This is just the beginning. The vision for Medinova includes:

  1. Full prescription scan and auto-detection of medicine schedules
  2. Wearable integration (like Fitbit or Apple Health) for proactive alerts based on vitals
  3. Expanding the pill identification engine using actual ML model training and third-party APIs
  4. Adding regional language support and audio-based interactions for accessibility in diverse communities
  5. Creating a care team dashboard for doctors or family to oversee patients remotely
  6. Exploring predictive health alerts, where Medinova learns from patterns to prevent missed doses or flag unusual symptoms

Ultimately, Medinova is not just a med tracker — it’s a mission to bring compassionate technology to people who need it most.

Built With

Share this project:

Updates