Kevin Carlberg, PhD

Venture Partner

Dr. Kevin Carlberg is a Venture Partner at Kaleida Capital, contributing expertise at the intersection of applied AI systems, contextual intelligence, and wearable/spatial computing. He was previously Director of AI Research at Meta, where he led multidisciplinary teams across AI, HCI, engineering, and design to develop novel AI and simulation technologies for next-generation wearable and mixed-reality platforms, including the CTRL Labs acquisition through Reality Labs Research. Prior to Meta, Kevin spent over eight years at Sandia National Laboratories, where he initiated and led research programs advancing AI-driven model reduction, large-scale uncertainty quantification, and real-time physics simulation for high-consequence national security applications.

Dr. Carlberg holds a PhD and MS in Aeronautics and Astronautics from Stanford University and a BS in Mechanical Engineering from Washington University in St. Louis. He also serves as an Affiliate Associate Professor of Applied Mathematics and Mechanical Engineering at the University of Washington. At Kaleida Capital, Kevin works closely with the investment team on NeuroAI and applied AI opportunities, while also building his own company at the intersection of embodied AI and wearable computing.

    • Director of AI Research Science at Meta Reality Labs Research | Built and led a 100+ member team spanning AI, HCI, engineering, design, and UX to develop physical AI and simulation technologies for Meta's wearable and mixed-reality platforms, including work spanning the CTRL Labs acquisition 

    • Distinguished Technical Staff - Sandia National Laboratories | Initiated and led research programs applying AI-driven model reduction and large-scale uncertainty quantification to enable extreme-scale physics simulations to run in near real time for high-consequence national security applications 

    • President Harry S. Truman Fellow at Sandia National Laboratories | One of the nation's most prestigious postdoctoral fellowships in national security science and engineering. 

    • Speaking Engagements | Frequent keynote, plenary, and invited speaker across leading AI, computational science, and applied mathematics forums, with 4 keynotes, 7 plenary lectures, and 35+ invited talks at institutions including MIT, UC Berkeley, Stanford, Cornell, NASA Ames, Pixar, and The Boeing Company, as well as AAAI, ICLR, and ICERM. 

    • Conference Panelist | Panelist at the ICLR 2024 Workshop on AI for Differential Equations in Science alongside Max Welling and Shirley Ho, representing the applied physics and scientific machine learning community.

    • Deep Learning & Data-Driven Dynamics 

      • Lee, K., & Carlberg, K. (2020). Model Reduction of Dynamical Systems on Nonlinear Manifolds Using Deep Convolutional Autoencoders. Journal of Computational Physics, 404, 108973. https://doi.org/10.1016/j.jcp.2019.108973

      • Carlberg, K., Jameson, A., Kochenderfer, M. J., et al. (2019). Recovering Missing CFD Data for High-Order Discretizations Using Deep Neural Networks and Dynamics Learning. Journal of Computational Physics, 395, 105–124. https://doi.org/10.1016/j.jcp.2019.05.041

      • Lee, K., & Carlberg, K. (2021). Deep Conservation: A Latent-Dynamics Model for Exact Satisfaction of Physical Conservation Laws. Proceedings of the AAAI Conference on Artificial Intelligence, 35, 277–285. https://doi.org/10.1609/aaai.v35i1.16102

    • Uncertainty Quantification & Hemodynamic Modeling 

      • Guzzetti, S., Alvarez, A., Blanco, P., Carlberg, K., & Veneziani, A. (2019). Propagating Uncertainties in Large-Scale Hemodynamics Models via Network Uncertainty Quantification and Reduced-Order Modeling. Computer Methods in Applied Mechanics and Engineering, 358. https://doi.org/10.1016/j.cma.2019.112626 

    • Nonlinear Model Reduction & Projection Methods 

      • Carlberg, K., Cortial, J., Amsallem, D., et al. (2011). The GNAT Nonlinear Model Reduction Method and Its Application to Fluid Dynamics Problems. 6th AIAA Theoretical Fluid Mechanics Conference (AIAA 2011-3112). https://doi.org/10.2514/6.2011-3112

      • Carlberg, K., Farhat, C., Cortial, J., & Amsallem, D. (2013). The GNAT Method for Nonlinear Model Reduction: Effective Implementation and Application to Computational Fluid Dynamics and Turbulent Flows. Journal of Computational Physics, 242, 623–647. https://doi.org/10.1016/j.jcp.2013.02.028

      • Choi, Y., & Carlberg, K. (2019). Space–Time Least-Squares Petrov–Galerkin Projection for Nonlinear Model Reduction. SIAM Journal on Scientific Computing, 41, A26–A58. https://doi.org/10.1137/17M1120531

    • Structure-Preserving & Conservative Model Reduction 

      • Carlberg, K., Tuminaro, R., & Boggs, P. (2012). Efficient Structure-Preserving Model Reduction for Nonlinear Mechanical Systems with Application to Structural Dynamics. AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference. https://doi.org/10.2514/6.2012-1969

      • Peng, L., & Carlberg, K. (2017). Structure-Preserving Model Reduction for Marginally Stable LTI Systems. arXiv. https://doi.org/10.48550/arXiv.1704.04009

      • Carlberg, K., Choi, Y., & Sargsyan, S. (2018). Conservative Model Reduction for Finite-Volume Models. Journal of Computational Physics, 371, 280–314. https://doi.org/10.1016/j.jcp.2018.05.019

    • Domains: Applied AI Systems, Physical AI, Wearable & Spatial Computing, Real-Time Physics Simulation, Contextual Intelligence

    • Strengths: Multidisciplinary research leadership, AI-driven model reduction, uncertainty quantification, deep-tech commercialization, embodied AI

    • Networks & Memberships: University of Washington (Affiliate Associate Professor), Stanford University (PhD, Aeronautics and Astronautics), SIAM (Society for Industrial and Applied Mathematics), ICERM (Institute for Computational and Experimental Research in Mathematics