Abubakarr Jaye

About

Abubakarr Jaye is an Applied Scientist at Microsoft working on large language model reasoning and agentic systems. His research focuses on building reliable evaluation frameworks for LLM agents, synthetic dataset generation, and improving reasoning robustness through multi-agent and verification-based methods. He is particularly interested in scalable evaluation, dynamic prompt engineering, and autonomous systems that learn from structured feedback.