Bilkent University · Computer Engineering
I like taking an idea all the way to a working product. I start with a clear question, build the first version myself, test it against real data and keep refining it until someone can actually use it. Most of my recent work sits where finance and software meet.
Technologies I work with
01Projects
From market models to hardware: things I built end to end
A system that runs subscription-based coding agents together on one git repository. An orchestrator plans the work; worker agents pick up tasks, lock the files they edit so they never collide, and merge into main once the tests pass. I built the web dashboard interface and the Codex CLI integration.
A local research workbench with a separate backtest engine, next-bar execution, explicit transaction costs and strategies written in Python. Data lives in SQLite; 78 unit tests pass.
A multi-journal platform that runs the whole editorial process, from submission and double-blind review to decision and publication. It supports Crossref DOIs, ORCID and OAI-PMH, includes a DergiPark archive migration tool, and rests on a security layer built on Argon2id, two-factor authentication and a strict CSP.
A browser-based trading journal that flags trades taken outside the plan, redraws the account without them and puts a number on the gap. Includes R-multiple statistics, Monte Carlo scenarios and drawdown analysis.
A study comparing Kalman and wavelet-based hedge estimators on the Visa/Mastercard pair. My own methodology review found two look-ahead leaks, which I fixed and locked down with regression tests.
A trilingual site I built with Next.js for a wedding and events company in Balıkesir. The business manages its blog and gallery from its own panel without touching code; the site includes service pages, a quote request form and 18 blog posts.
A game our five-person team built with LibGDX for Bilkent’s CS102 course. Each of its six levels ends in a boss fight themed after one of the Seven Deadly Sins. I owned the UI/UX, screen navigation and Firebase integration.
A Raspberry Pi computer vision system that counts the people in a building and passes that number to emergency teams. A companion Flutter app shows assembly points and lets survivors send requests.
A home-scale soilless vertical farming unit that automatically provides the right conditions for the selected plant, monitored and controlled remotely through a Flutter app.
02About Me
My education, skills and the questions that drive me
I study Computer Engineering at Bilkent University in an English-medium program (GPA 3.52 / 4.00, expected graduation 2029). Since high school I have been doing the same thing: taking a problem and turning it into a system that actually works. The tools change (Raspberry Pi and Arduino for hardware, Flutter for mobile, Java for a desktop game, Python for data), but the approach stays the same: see it through to the end.
Over the past year my focus has moved toward financial markets and time-series research. I am comparing Kalman and wavelet-based hedge estimators on Visa/Mastercard; I reviewed that code myself and closed its look-ahead leaks with regression tests. With the same approach I built a BTC/USDT backtesting workbench and a trading journal (TradeTracker) that measures what following your own rules is worth.
My goal is to join a team where finance and technology meet, work with real capital and real data, and learn from people who do this every day.
Board member, Bilkent IBM Z Student Club
03Contact
Get in touch with me
If you would like to talk about an internship, a collaboration or one of my projects, fill in the form or email me directly.
+90 537 277 53 14
