Safe Rl Prompting for Llms for Solving Data Management Challenges
- 2023
- INFORMS 2023
Dr. Alexander (Sasha) Zadorojniy is a research scientist at IBM Research – Israel, with over 14 years of experience in developing efficient algorithms for real-world optimization and decision-making problems. He holds a Ph.D. in electrical engineering from Tel Aviv University and B.Sc. and M.Sc. degrees from the Technion – Israel Institute of Technology. During his doctoral work, Sasha developed a novel strongly polynomial algorithm for controlled queues, demonstrating theoretical rigor and numerical stability—results that were published in Mathematics of Operations Research.
His expertise spans Reinforcement Learning, Markov Decision Processes (MDPs), and mathematical optimization. His recent work on risk-aware MDPs (Extended MDPs, or EMRDPs), co-authored with Orit Davidovich and Takayuki Osogami, was accepted to the ICAPS 2025 main track, and a prior paper appeared in the IJCAI 2023 main track. Sasha’s algorithms have been successfully implemented in real-world settings, including the optimization of wastewater treatment plants (WWTPs).
He is a recipient of the Mehrez Prize and has been recognized as a finalist for the Wagner, Nicholson, IAAA, and GYSS awards. Prior to joining IBM, he spent a decade at Intel, leading design efforts in high-performance processors with a focus on logic design and complex clock networks. Sasha is also a two-time IBM Master Inventor.
Currently, Sasha is driving research on risk‑aware Extended Markov Ratio Decision Processes (EMRDP), focusing on scalable, GPU‑enabled parallel algorithms.