Computer Engineer · AI & Web Application Developer
I build software that ships. My work spans computer vision, deep learning and modern web applications — from a TÜBİTAK-supported physiotherapy analysis system to LSTM-based financial forecasting models and live web products used by real users.
| Project | Description | Stack |
|---|---|---|
| PomoStat | Focus & habit tracking app that combines the Pomodoro technique with an analytics dashboard, making time usage visible through charts. | React · Vite · Supabase · D3.js |
| Pratik İngilizce | Free vocabulary & grammar platform for academic English exams (YDS/YÖKDİL). Measures text coverage with a 32-question placement test, then closes the gap through 24 frequency-ordered word pools (1574 words) and Leitner spaced repetition. No account, no ads — progress stays on the device. | Astro v5 · SSG · localStorage |
| UseToolSuite | Privacy-first productivity platform — a curated set of tools that run 100% in the browser, with no data ever leaving the user's device. | JavaScript · Modern UI · SEO |
-
AI-Based Real-Time Physiotherapy Exercise Analysis (TÜBİTAK 2209-A supported) Detects joint landmarks with MediaPipe and evaluates exercise form in real time, giving instant feedback to the patient and remote progress tracking to the physiotherapist.
PythonComputer VisionMediaPipe -
NTradeWeb — Stock Price Prediction (B.Sc. graduation project) Predicts the next trading day's closing price for Borsa İstanbul equities using LSTM-based deep learning models, served through a Streamlit interface.
PythonTensorFlowLSTMStreamlit
Languages: Python · Java · JavaScript · SQL AI / Data: TensorFlow · MediaPipe · LSTM & time-series modeling · OpenCV Web: React · Vite · Astro · HTML5 · CSS3 · Supabase · D3.js Tools: Git · Streamlit · REST APIs
- Growing PomoStat, Pratik İngilizce and UseToolSuite as real products — performance, SEO and usability
- Privacy-first product design: keeping user data on the client side wherever possible
- Content-driven static architectures with build-time validation scripts that block a release when data integrity checks fail
- Deepening my backend and system design skills alongside applied machine learning
- Turn ideas into scalable, production-grade software systems
- Build products that solve real problems for real users, not just demos
- Keep pushing at the intersection of artificial intelligence and web engineering