Ethan Lew
I build AI and design optimization methods that speed up real engineering work. Lately that means design automation and AI for design engineering, so teams can take on more complex designs and finish them faster.
I lead customer engineering at P-1 AI, where we're building an AI engineer for the physical world. Before that, I spent five years at Galois working on formal methods and autonomy.
I'm most interested in AI research and formal methods as they apply to design, especially getting AI to produce designs that engineers can check and trust.
Papers
- Fast Koopman Surrogate Falsification Using Linear Relaxations and Weights ATVA 2025
- Falsification using Reachability of Surrogate Koopman Models HSCC 2024
- Learning nonparametric ordinary differential equations from noisy data J. Comput. Phys 2024
- AutoKoopman: A Toolbox for Automated System Identification via Koopman Operator Linearization ATVA 2023
- Leveraging Manifold Learning and Relationship Equity Management for Symbiotic Explainable Artificial Intelligence AHFE 2023
- Reachability of Koopman Linearized Systems Using Random Fourier Feature Observables and Polynomial Zonotope Refinement CAV 2022
Writing
- Trollfish∶ A Trolling Chess Bot 2024
- Product Line Engineering Feature Selection with Chocosolver 2024
- Shaping Tomorrow’s Innovators 2023
- Driving to a Secure Future: Demonstrating a Vehicle That Thwarts Cyberattacks 2022
- Regression and Classification with Kernels 2022
- Head-to-Head Races: How Galois Placed First in the GRAIC 2022
- Real-time Robotics Control in the Lean Language 2021