I’m an Assistant Professor at TU Delft, where I lead the Tapri Lab within the Department of Intelligent Systems. My work explores the foundations of intelligence by developing multimodal generative models that enable AI systems to perceive, reason, and act in dynamic environments.
I am particularly interested in how navigating social complexities shaped the evolution of human intelligence and what this implies for embodied AI systems. Consequently, I leverage multi-agent interactions across physical and virtual environments as a framework for modeling continually adaptive behaviors. To this end, I draw upon insights from cognitive science and social psychology to inform my computational techniques. Learn more about my research through my publications.
As part of the Hybrid Intelligence Center, I co-lead the Robotic Surgery case study, developing coupled human-agent systems for microsurgery in collaboration with Erasmus Medical Center, Rotterdam. I am also a visiting researcher at the DIS Group, CWI Amsterdam, where I develop ML techniques for facilitating remote immersive interactions.
If you’d like to get in touch, send me a note!
About me
Software engineer, game developer, animator, and UX designer turned computer scientist—exploring intelligence, interaction, and everything in between.
I’ve always been fascinated by how people communicate, adapt, and make sense of the world together. This curiosity led me through diverse roles spanning engineering, design, and computer vision research before leading me to multimodal machine learning and behavior modeling. Below is an overview of my journey.
Outside of research, I have a deep love for storytelling, design, and immersive experiences—whether through animation, games, or exploring new places on my bike.
- Ph.D. Computer Science, highest honors ∙ TU Delft, The Netherlands ∙ 2023
- Master of Entertainment Technology ∙ Carnegie Mellon University, USA ∙ 2013
- Bachelor of Engineering (Information Technology) ∙ University of Mumbai, India ∙ 2010
After seven years in the US, I moved across the pond to the Netherlands in Sep 2018 to pursue my Ph.D. at TU Delft and explore a different research culture. My thesis made cross-disciplinary contributions towards bringing artificial social intelligence research into real-world settings, focusing on multiple facets including data acquisition and synthesis, modeling and forecasting of behavioral dynamics, and perception of embodied social behavior in multiparty interactions. During my Ph.D., I also spent a summer at Microsoft Research Cambridge, generating more realistic synthetic human data for downstream vision tasks.
After my Ph.D., I transitioned into my current role as a tenure-track Assistant Professor at TU Delft in Mar 2023, where I combine my passions for 3D behavior generation, multimodal representation learning, and social psychology, with downstream applications in healthcare and education.
Prior to my Ph.D., I was a Senior Research Engineer at Carnegie Mellon University, working in LP Morency’s MultiComp Lab. I developed the CMU Multisense platform for sychronized multimodal behavior capture and modeling, including realtime integration of multimodal machine learning models. I was an early adopter of Microsoft’s Platform for Situated Intelligence (\psi), providing design feedback and later presenting my findings at Microsoft Research. During this time, I also collaborated on projects developing multimodal techniques for emotion-aware interactive dialog systems, curiosity sensing, and driver distraction detection.
My startup phase! I was one of the founding engineers at ProductionPro, a New York-based startup building a real-time collaboration platform for film, TV, and theatre productions. As Lead Client Developer and UX Designer, I architected and implemented the core client-side systems for the flagship iOS platform, which later saw adoption on Netflix and Disney productions. I also defined and led the company’s UX design process, ensuring a seamless experience for creative teams in high-paced production environments.
One of the most fun jobs I’ve had! At Disney Research Pittsburgh, I had the opportunity to develop computer vision approaches for real-world challenges in theme park environments. I developed scalable infrastructure for distributed training and inference of object detection models using data captured across different Disney park locations. One of the more unique projects I worked on was detecting and tracking giraffe feeding behavior. I have fond memories of these good old days when deformable parts-based models were still state-of-the-art for object detection.
Inspired by Randy Pausch’s Last Lecture on Really Achieving Your Childhood Dreams, I pursued the Master of Entertainment Technology (MET)—a unique interdisciplinary degree that Pausch co-founded as a joint venture between Carnegie Mellon University’s School of Computer Science and the College of Fine Arts. I spent a semester with Microsoft’s User Experience and Creative Services team, developing an interactive projection-mapped installation for the Microsoft Store. I spent the second year of program on an integrated co-op, where I worked as a New Technology Analyst with Walt Disney World’s Next Generation Experience team. This involved developing methods for automated color-retargeting of guest media captured on attractions and integrating high-speed video capture into Disney’s existing media pipeline.
During the last two years of my Bachelor’s, I worked at IIT Bombay as part of Project OSCAR, a government-funded initiative for expanding distance education through open-source animated content. I developed immersive virtual labs using Blender 3D for interactive science education in rural India and contributed to MOOCs. This experience shaped my understanding of how privilege is multidimensional—growing up in a Tier-1 city like Mumbai provided me access to resources many in rural areas of both India and the developed world lack. It also sparked my passion for the transformative potential of 3D immersive media, game engines, and open source software.
After graduating, I worked at Hungama Digital Media in Mumbai, developing games, speech-driven iOS apps, and gesture-based installations for clients like Cartoon Network, Axis Bank, and Yash Raj Films, laying the foundation for my future work in interactive media and AI-driven experiences.