Show, don't tell.
Five products I built in 2026; four are open source, with receipts you can check. Everything else is below, grouped by category.
Checks AI-written diffs, masks the secrets it finds and gates the merge — models optional.
Follows what a conversation is about — and can answer “uncertain” instead of guessing.
Answers predictive questions straight from related tables — leakage-tested and explained.
Catches the agent regressions a 200 OK hides — down to the step where they begin.
An astrology companion whose AI interprets charts computed with the Swiss Ephemeris.
Özgür Işık DamarSenior Software EngineerI'm Özgür Işık, a senior software engineer based in Türkiye. I build production systems across mobile, web, data, AI and e-commerce. In 2026 I shipped five products: AgentTwin, ContextLens, RealPath, Membrane AI and Universal AI Astro. My current focus is Agentic AI, Search Engines, E-export and Cross-Border Logistics. I build for real users, real operations and real scale.
Now
My demo refund agent returned 200, claimed SUCCESS and refunded twice. How I catch that: tool twins that record every call, and faults that split the reply from the state.
My locked test set ran twice: one field hid 1,151 questions from the headline score. How I fixed the data, proved the model never moved, and published both runs.
RealPath's leakage claim is a pytest: delete everything after the anchor, rebuild, diff. What it caught when I broke the cutoff, and two leaks it still misses.
Practical choices, not résumé padding. The list is short on purpose.
I take on a small number of engagements each quarter: full-stack builds, mobile rescues, data-flow architecture and AI plumbing. If your problem sounds interesting, write.