Summary
Anurag Koul is a Ph.D.-trained researcher-engineer with 11 years of experience advancing reinforcement learning and reasoning for code generation, currently driving R&D as a Member of Engineering in New York. His work spans foundational RL (world modeling, state abstraction, hierarchical and offline methods), safety and multi-agent systems, and practical RL-for-LLM approaches that improve code-editing and post-training reasoning. At AWS and Microsoft he built evaluation benchmarks, RL training pipelines using code-execution feedback, and multi-LLM routing strategies to speed low-latency inference—bridging research rigor with production constraints. He combines hands-on systems development from earlier software roles with deep academic expertise, and uniquely focuses on using RL-driven planning and in-context trajectory imagination to boost LLM reasoning for complex code and math tasks.
10 years of coding experience
10 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at Oregon State University
Bachelor of Engineering (BEng) Computer Science, Bachelor of Engineering (BEng) Computer Science at University of Mumbai
English, Hindi