DEVLOG #1

Built a web app that connects to smart glasses (Raspberry Pi + camera) to detect eye diseases in real-time.

What I Made

Veya Web App - Next.js platform that connects to Veya Glasses via WiFi for instant eye disease screening.

Features:

• Connect to Raspberry Pi via IP address
• Capture eye image through camera
• AI detects cataracts, conjunctivitis, pterygi
• 85%+ accuracy with confidence scores
• Scan history (last 10 scans)
• Bilingual (English/Russian)
• Dark/Light theme
• Privacy-first (all processing on device, no cloud)

Tech: Next.js 16, TypeScript, Tailwind CSS, TensorFlow Lite on Raspberry Pi

Live demo: https://veya-web-zeta.vercel.app

Challenges

1. Real-time disease detection pipeline
   Need to capture eye image → send to TensorFlow Lite model → return diagnosis in <3 seconds. Built demo mode first to test UI,
   now integrating real Flask API on Pi.

2. WiFi connection flow
   Users connect web app to glasses over local network. Added IP validation, error handling for timeouts/wrong IPs.

3. Privacy requirements
   Medical data can’t go to cloud. All AI processing happens locally on Raspberry Pi.

What’s Next

• Connect real Flask API from Raspberry Pi
• Test with actual TensorFlow Lite model
• Add scan history persistence
• Export results as PDF

Why It Matters

Eye diseases affect millions but professional e $50 hardware (Pi + camera) instead of $15,000medical devices. Built for rural areas in Kazakhstan where eye care access is limited.