Ranjit Kumaresan
My current work focuses on the security of LLM-based systems — red-teaming agentic AI, security benchmarking, and building agent harnesses. I’m a Research Scientist at Visa Research, on the GATE team (GenAI Trust & Efficiency).
Earlier, my research spanned the design and analysis of secure systems more broadly: LLM watermarking via pseudorandom error-correcting codes, MEV in DeFi, universal payment channels, and the foundations of secure multiparty computation. Before Visa I was at Microsoft Research, held postdoc positions at MIT CSAIL and Technion, and earned my Ph.D. from the University of Maryland.
I’ve been fortunate to have great mentors and talented interns.