Hi, I'm Dogukan. I'm an applied AI engineer and researcher obsessed with learning infinitely, exploring and building many things. My research and curiosity focuses on frontier RL x inference infrastructure, GPU/TPU kernels, agentic AI infrastructure (retrieval, memory, knowledge bases) and distributed systems.

all my post formats are inspired by Lilian Weng, Thinking Machines, Henry Ko.

What I enjoy most:

* Learning infinitely

* New places and people

* Spending time in nature, long walks and new trails


You can find me on X (DMs Open), GitHub

Research & Thoughts

Defeating Nondeterminism in LLM Inference: Reproducing Batch-Invariant Ops (RMSNorm & Tiled Matrix Multiplication) in JAX

Streaming deepagents and task delegation with real-time output

Energetics of Allosteric Communication in Ubiquitin Revealed by Hybrid MCTS-Langevin Simulations