ROHIT M | AI Researcher

AI Researcher @ SRM Institute of Science and Technology

me.png

SRM Institute of Science and Technology

Chennai, India

I am ROHIT M (also known online as PoppingPixel), a fourth-year Computer Science and Engineering student at SRM Institute of Science and Technology, conducting research under the supervision of Dr. Pushpalatha M. My research focuses on Multi-Agent Reinforcement Learning (MARL) for quantitative finance and Vision Transformers (ViTs) for medical imaging diagnostics.

To translate these foundations into practice, I developed QuantAlpha, a multi-agent LLM architecture to navigate high-frequency trading dynamics on the National Stock Exchange (NSE). In medical imaging, I engineered self-supervised contrastive learning (SimCLR) pipelines integrated with ViTs for automated brain tumor detection.

Beyond research, I completed the McKinsey Forward program and participated in VC Lab Cohort 21. I collaborate on deep learning projects with Pradeep, Tamilselvan, and Charan. I am currently preparing for direct PhD applications at Stanford University (Dec 2026) and the WSAI Summer Internship at IIT Madras.

news

Apr 26, 2026 Excited to announce that our research paper titled “Deterministic Credit Assignment for Multi-Agent Portfolio Management via a Provenance DAG” has been successfully submitted to an IEEE conference as part of academic track 21CSP302L! This work introduces a novel framework mapping Directed Acyclic Graphs (DAGs) to resolve cooperative agent reward tracking in chaotic market environments.
Feb 21, 2026 Attended the India AI Impact Summit 2026 at Bharat Mandapam, New Delhi (February 16 – February 20, 2026). It was a phenomenal opportunity to interface with top-tier AI researchers, system architects, and national infrastructure planners driving the next generation of scalable AI solutions! :sparkles:

latest posts

selected publications

  1. ICML-W
    quantalpha_returns.png
    Deterministic Credit Assignment for Multi-Agent Portfolio Management via a Provenance DAG
    M Rohit
    May 2026
    Submitted to ICML 2026 Workshop on Reinforcement Learning from World Feedback (RLxF) (Submission #130)