Ankit Shah

Ankit Shah is a Postdoctoral Researcher at Brown University's Department of Computer Science, collaborating with Profs. Stefanie Tellex, George Konidaris, and Michael Littman. His research focuses on developing computational models that enable domain-experts to train autonomous agents as human apprentices.

• PhD in Autonomous Systems (2021), MIT
• SM in Aeronautics and Astronautics (2016), MIT
• B.Tech in Aerospace Engineering (2013), IIT Bombay

His research interests center on Robotics, Machine Learning, Temporal Logic, and Human-Machine Interaction. He specializes in enabling domain-experts to directly train autonomous systems through demonstrations and criticism, emphasizing transparency and non-Markovian task specifications.

His publications (2024-2020) demonstrate expertise in Temporal Logic, Bayesian Inference, Interactive Training, and Safe Motion Planning for robotics. Recent work explores large language models for command translation and safety constraints in LLM-driven agents.

At Brown University, Shah actively contributes to robotics research, focusing on model transparency, manipulation planning, and specification learning through examples.