SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Physical Superintelligence is building AI systems to discover new physics at scale, with roots at Google, NVIDIA, Harvard, Meta, MIT, Princeton, Oxford, Johns Hopkins, Cambridge, and the Perimeter Institute. The company is seeking applied physicists, engineers, and computational scientists to apply AI to real engineering systems including thermal and fluid systems, electrical power distribution, and coupled models. The mission is to discover and commercialize transformative physics breakthroughs using artificial superintelligence, safely and verifiably.
In this role, you will model real physical and engineering systems and validate models against solvers and measured data. You will use learned models and simulators together, deciding where solver fidelity is required versus where learned models suffice, and use them to search design spaces. You will specialize deeply in one domain—fluids and heat transfer, thermodynamics and energy conversion, electrical and power systems, electrochemistry, solid mechanics and materials, transport and environmental physics, or chemical and process engineering. Each engagement solves actual customer problems, ships something they can run, and builds reusable capabilities.
The company emphasizes high technical standards, full ownership from spec to ship to on-call, writing contracts before logic, testing against real systems, and favoring simple designs that ship. Development is AI-native, using agentic coding tools daily, specs legible to humans and agents, and leading with leverage.
REQUIREMENTS:
- Three or more years building physics-based models of real engineering systems (simulations, reduced-order models, or first-principles analyses). Must understand which assumptions, boundary conditions, and mesh/discretization choices drive results, have checked results against experiments or measured data, and can articulate discrepancies.
- Three or more years applying scientific machine learning to real physics or engineering problems (not synthetic benchmarks). Must have used, validated, and deployed learned models such as surrogates, neural operators, or physics-informed networks, and understand out-of-distribution behavior, extrapolation failures, detection methods, and fallback strategies.
- Deep expertise in one engineering domain with knowledge of relevant codes and standards: fluid dynamics and heat transfer (CFD, conjugate heat transfer, multiphase flow, convection); thermodynamics and energy conversion (refrigeration, power cycles, heat exchangers, combustion, engines, turbines, HVAC); electrical and power systems (power flow, interconnection, protection, power electronics, electromagnetics); electrochemistry and energy storage (batteries, fuel cells, corrosion); solid mechanics and materials (FEA, vibration, fatigue, degradation); transport and environmental physics (atmospheric dispersion, hydrology, subsurface flow, fire); or chemical and process engineering (reaction kinetics, process simulation, working fluids).
NICE TO HAVE:
- PhD or equivalent research record in engineering, applied physics, computational science, or machine learning for science.
- Model deployed in a system others depend on (behind API, versioned, with tests).
- Training runs designed and debugged in PyTorch or JAX; experience with neural operators, physics-informed networks, or uncertainty quantification.
- Optimization over design spaces (mixed-integer, derivative-free, or Bayesian search).
- Peer-reviewed publications in computational physics, engineering, or ML for science.
