Dogukan Uraz Tuna

I'm a researcher exploring the frontiers of reasoning, multi-agent systems, program synthesis and scaling laws. My goal is to push the boundaries of what's possible at the intersection of AI and superhuman scientific reasoning by building extraordinarily productive innovator-agent systems for science and code generation. Build systems that accelerates AI research and development in an autonomous way, and empower researchers and developers to focus on the most important problems or their own AI products.

Currently, I'm focused on recursive self-improvement (RSI), compute-optimal reward/verifiers for autonomous reasoning agents, infinite-context and program synthesis. I see these areas as critical to enabling agent systems capable of true innovation in science.

I believe that advancing scientific multi-step reasoning—and integrating it with agentic systems—is a key stepping stone toward superhuman scientific reasoning. To that end, I'm experimenting with novel ideas, building prototypes, and embracing open source as the best way to share knowledge.