Meet the Deccan AI team @ IROS 2026 | Booth 1114
Session and Timing Details
We'll be at IROS, Pittsburgh, September 27–October 1.
With over 1,900 papers and 170+ exhibitors this year, it's shaping up to be the biggest robotics gathering of the year, and squarely the room where the data problem behind all of this gets talked about.
On September 29 at 3:00 PM, our robotics lead Narendar Sriram will be speaking on "Data: The Unsolved Enigma of Physical AI," on why robot foundation models don't have an internet-scale corpus to learn from the way language models do, and what that means for how you think about collection modalities, demonstration diversity, and data curation.
We'll also be running lightning talks at Booth 1114 throughout the conference, and giving away an NVIDIA Orin Nano plus a bunch of other goodies. Animesh Maheshwari, Narendar, Amanda Brackbill, Jagdeep Sahni, and Samir Janveja will be at the booth all week. Come find us if you're working on physical AI and want to talk about the data problem, catch the talk, or just enter the giveaway.
Meet the speakers:
Narendar Sriram leads Robotics Machine Learning at Deccan AI. An alum of Carnegie Mellon's Robotics Institute, he brings 7 years of experience building learning-based robotic systems.
At Deccan AI, he's building the robotics data flywheel and working on world action models, VLAs, and the data infrastructure to create superaccurate robot foundation models.
Animesh Maheshwari is a Robotics researcher at Deccan AI, where he works on data and evaluation research spanning VLA models and world models. A graduate of the University of Texas at Dallas and its Intelligent Robotics and Vision Lab, he brings a background in robotics and multimodal AI research.
About Deccan AI
Deccan AI builds the post-training stack (RL, SFT, Evals) for frontier AI labs across physical, multimodal, and agentic AI. Our Physical AI team focuses on one of the hardest open problems in robotics: unlike language models, robot foundation models have no internet-scale corpus to learn from.