Soumya Rani Samineni

I am a second-year Ph.D. student at Arizona State University with 3+ years of professional experience and over 2 years of research experience in Artificial Intelligence (AI), Reinforcement Learning (RL), Large Language Models (LLMs), Machine Learning (ML), optimization, and robotics.

I worked as an ML Research Engineer at Quantiphi, Bangalore, where I applied reinforcement learning to workforce optimization. I was a Research Fellow at Microsoft Research India, focusing on reinforcement learning algorithms for energy grids. I also contributed as an AI Engineer at AI Labs, Hyderabad, developing a quadrupedal controller inspired by MIT Cheetah’s impedance control and building object-detection models.

I completed my Master’s in Computer Science and Engineering at the Department of Computer Science and Automation, IISc Bangalore, advised by Prof. Shishir Kolathaya and Prof. Shalabh Bhatnagar. My master’s thesis, Policy Search using Dynamic Mirror Descent for Off-policy Reinforcement Learning, was funded by the Robert Bosch Center for Cyber-Physical Systems (RBCCPS). During my time in the Stochastic Systems Lab and the Stochastic Robotics Lab, I studied reinforcement learning for robotics and stochastic approximation.

Prior to IISc, I served as an Assistant Executive Engineer (Civil) for the Government of Telangana and hold a Bachelor's degree in Civil Engineering from National Institute of Technology, Warangal (NITW).

Email  /  Resume  /  Google Scholar  /  Twitter  /  Github

I’m interested in Artificial Intelligence, with a focus on reinforcement learning, large language models, machine learning, optimization, and deep learning. My research spans reinforcement learning–based post-training methods for large language models, as well as studying their reasoning and planning limitations.