@Ulagat

Veya - AI-powered smart glasses that detect eye diseases A wearable device using Raspberry Pi and computer vision to screen for cataracts, conjunctivitis, and pterygium. Captures eye images, runs a TensorFlow Lite CNN model locally, and sends diagnosis + confidence score to a mobile app via WiFi. Built for underserved communities in Central Asia where ophthalmologists are scarce. Under $350 vs $10k+ professional screening devices. All processing happens on-device (no cloud, privacy-first).

DIY CNC Plotter with 28BYJ-48 Stepper Motors A 3-axis drawing machine that converts digital vector images into physical drawings. Built around three cheap 28BYJ-48 stepper motors, an Arduino Nano, and GRBL firmware. Takes SVG files from Inkscape, converts them to G-code, and plots them with a pen. The whole thing costs under $20 in parts and can draw anything you can trace in vector format — logos, portraits, technical drawings, whatever. Hardware is dead simple: steppers for X/Y movement + one for pen lift/drop, ULN2003 driver boards, and standard Arduino. No custom PCBs, no expensive parts. Software stack is Arduino IDE → GRBL library → Inkscape for prep → any G-code sender. It's a learning project that actually produces something useful — bridges embedded programming, motion control, and CNC workflow without needing a full machine shop.

I'm building KozbenSal — smart glasses that let paralyzed people control computers and draw using only eye movements. The Problem: People with motor disabilities (ALS, cerebral palsy, spinal injuries) can't use keyboards or mice. Existing eye-tracking solutions cost $10,000+. The Solution: KozbenSal uses a Raspberry Pi + camera to track eye movements in real-time. Users can: - Control mouse cursor by looking - Draw digital art with eyes - Click by blinking - Navigate applications hands-free Tech Stack: - Hardware: Raspberry Pi 4, Pi Camera v2, 3D-printed frame (~$150 total) - Software: OpenCV + MediaPipe for eye tracking - Latency: <100ms response time - Calibration: 9-point personalized setup

I'm building KozbenSal — smart glasses that let paralyzed people control computers and draw using only eye movements. The Problem: People with motor disabilities (ALS, cerebral palsy, spinal injuries) can't use keyboards or mice. Existing eye-tracking solutions cost $10,000+. The Solution: KozbenSal uses a Raspberry Pi + camera to track eye movements in real-time. Users can: - Control mouse cursor by looking - Draw digital art with eyes - Click by blinking - Navigate applications hands-free Tech Stack: - Hardware: Raspberry Pi 4, Pi Camera v2, 3D-printed frame (~$150 total) - Software: OpenCV + MediaPipe for eye tracking - Latency: <100ms response time - Calibration: 9-point personalized setup

I'm building Veya — affordable smart glasses that detect eye diseases using AI, bringing diagnostic screening to communities without access to ophthalmologists. The Problem: Cataracts cause 51% of world blindness (WHO), yet professional screening equipment costs $10,000+. In rural Kazakhstan and similar regions, people travel hours to reach eye clinics, often detecting problems too late. The Solution: Veya uses a Raspberry Pi 4 + camera module to capture eye images and run a TensorFlow Lite CNN model that detects three diseases in under 3 seconds: - Cataract (leading cause of blindness) - Conjunctivitis (highly contagious, needs early treatment) - Pterygium (progressive condition) All processing happens locally (no cloud needed), results display on a mobile app via WiFi with confidence scores + doctor recommendations.

Veya - AI-powered smart glasses that detect eye diseases A wearable device using Raspberry Pi and computer vision to screen for cataracts, conjunctivitis, and pterygium. Captures eye images, runs a TensorFlow Lite CNN model locally, and sends diagnosis + confidence score to a mobile app via WiFi. Built for underserved communities in Central Asia where ophthalmologists are scarce. Under $350 vs $10k+ professional screening devices. All processing happens on-device (no cloud, privacy-first).