À propos
I am a Research Intern at Microsoft Research, where I work with Akshay Nambi on understanding the capabilities and limitations of AI systems. My research focuses on evaluation methodologies, failure analysis, and controlled environments for studying agentic systems, computer-use agents, and conversational AI. I am particularly interested in developing diagnostic frameworks that move beyond measuring whether a system succeeds toward understanding why it succeeds or fails.
My recent work spans failure-centric evaluation of agentic systems, synthetic environments for training and evaluating computer-use agents, and multi-agent marketplaces for studying economic behavior, decision making, and interaction dynamics. More broadly, I am interested in AI evaluation, agentic systems, human-AI interaction, reliability, and the design of experimental environments for understanding AI behavior in real-world settings.
Prior to Microsoft Research, I was a Project Research Associate at CFILT, IIT Bombay, where I worked with Prof. Pushpak Bhattacharyya (opens in new tab) on conversational AI evaluation and speech technologies.