About
I’m a CS Master’s student at ETH Zurich, working on retrieval, reinforcement learning and how to make ML systems go faster. Right now I’m training LLMs to be Socratic tutors (the kind that help students think instead of just giving away the answer) using multi-stage RL (DPO, GRPO).
Before Zurich, I did my undergrad in Computer Engineering at METU in Ankara (3.91/4.0, 6th out of 286), and spent a semester at Saarland University / Max Planck Institute for Informatics through Erasmus.
Outside of research, I like building things. I’ve won a couple of hackathons, and I keep building stuff that I’m interested in.
What I work on
Language Model policy optimization (DPO, GRPO, PPO), AI for Education, Retrieval Systems, LLM alignment and evaluation, scalable ML, systems for ML.
Recent news
- Sep 2025 – Started my MSc at ETH Zurich.
- May 2025 – Published OpenEthics, where we evaluated how open-source LLMs handle ethical reasoning in Turkish and English.
- Jul 2025 – Graduated from METU (6th/286).