Dr Rabimba Karanjai

Dr Rabimba Karanjai leads the Institute’s work on Agentic AI: the question of what autonomous systems do once they hold real permissions rather than a sandbox. It is the frontier where the deployment gap is widest, because an agent that acts is an agent that can be wrong at speed and at scale.

He won the ACM SIGSOFT Distinguished Paper Award at AIware 2025 for Smartify, an agent-based framework that finds and repairs security vulnerabilities in smart contracts as they are written, reproducing a professional audit without waiting for one. His doctorate, Teaching AIs to Reason and Code, Confidentially, took the University of Houston’s Dan E. Wells Outstanding Dissertation Award for Fall 2025, setting out a multi-agent architecture in which reliability comes from how the system is arranged rather than from trusting any single model.

He was selected for the Google Cloud Research Innovators programme as the only student in his cohort. He is a Google Developer Expert in artificial intelligence and machine learning and a staff machine learning scientist at PayPal, and previously worked with Mozilla Research’s mixed reality team and IBM Research.