Medicine Scanner App

• 

  Splash Screen – Branding for Medicine Scanner App powered by Perplexity Sonar API.

• 

  Dashboard – Shows list of previously scanned medicines with a delete icon and “Scan a Medicine” button.

• 

  Scanning Medicine – Camera interface powered by ML Kit that scans text from medicine labels.

• 

  Scan Confirmation – Dialog prompts user to confirm whether the scanned text contains a medicine name.

• 

  Details Screen – Displays name, purpose, dosage frequency, side effects, and citations fetched via Sonar API.

• 

  Delete Medicine – Allows user to delete the scanned medicine entry stored locally.

• 

  Share Medicine Information – Demonstrates that scanned medicines information can be shared via communication apps

💡 Inspiration

Many people struggle to understand complex medicine labels — especially the elderly, those with visual impairments, or people managing multiple prescriptions. I wanted to create a tool that lets users instantly scan medicine names and get reliable, AI-generated details about them — without needing to Google or decipher medical jargon.

What I Learned

• Working with the Perplexity Sonar API to fetch AI-driven structured information
• Integrating ML Kit’s on-device text recognition in real-time
• Implementing Room DB for offline persistence in a clean architecture setup.
• Managing Jetpack Compose and state flows in a multi-screen architecture.
• Designing a smooth UX for scanning, validating, and displaying medicine data

🏗️ How I Built It

• I used Android Studio and built the app using Jetpack Compose and MVVM with Clean Architecture.
• Text recognition was implemented using ML Kit's TextRecognizer API.
• After confirming scanned text, I made a call to the Perplexity Sonar API to fetch structured medicine details.
• Data was stored in a local Room Database, ensuring the app works even without internet.
• Citations were shown as clickable links that open in Chrome.

🧱 Challenges I Faced

• Ensuring accurate and readable text extraction from varying medicine label fonts and angles
• Designing a simple and non-intrusive UX for scan confirmation
• Managing API data formatting to map correctly into local data models
• Handling edge cases like empty scans, or citation formatting issues.

Built With

• android-studio
• jetpack-compose
• kotlin
• retrofit
• roomdb
• sonar