Accelerated Physics Simulation Engineer - Agentic Computational Engineering

About the role

This role develops high-fidelity, accelerated physics simulation capabilities for AI-driven design optimization within the Agentic Computational Engineering team. You will architect and implement simulation kernels that exploit modern GPU architectures to reduce compute time for complex physical systems. The work centers on tightly coupled physics models where numerical precision and throughput directly influence design decision quality. You will partner with research scientists and software teams to integrate novel methods into scalable production pipelines. The role demands comfort with translating mathematical formulations into performant, verifiable code under aggressive timelines. You will own end-to-end simulation modules from requirements analysis through benchmarking and deployment. Communication clarity is essential as you explain intricate numerical behavior to both technical and non-technical stakeholders.

Key facts

What you'll do

Implement high-performance numerical solvers for partial differential equations using finite element, finite volume, and finite difference methods.

Design and integrate GPU-accelerated kernels with CUDA to achieve significant reductions in simulation wall-clock time.

Build surrogate models that approximate expensive physics simulations while preserving critical fidelity for rapid iteration.

Leverage large language models to generate, refactor, and validate simulation code, creating an AI-first development workflow.

Diagnose performance bottlenecks in existing simulation pipelines and apply optimization strategies targeting memory and compute efficiency.

Collaborate with domain experts to translate physical requirements into robust, testable simulation algorithms.

Establish automated testing and validation frameworks that ensure numerical accuracy and reproducibility across simulation variants.

Contribute to open-source performance libraries and internal tools that accelerate simulation development across the organization.

Analyze empirical timing and accuracy data to demonstrate quantitative improvements before and after optimization efforts.

Maintain documentation that captures design decisions, numerical methods, and usage patterns for long-term maintainability.

Partner with machine learning teams to align simulation outputs with downstream data-driven models and optimization objectives.

Drive the adoption of best practices in scientific computing, version control, and continuous integration specific to HPC workflows.

Evaluate emerging hardware capabilities and assess their suitability for accelerating core simulation kernels.

Champion code review processes that emphasize correctness, performance, and clarity to elevate team engineering standards.

Requirements

Hold a PhD in Computational Physics, Mechanical or Aerospace Engineering, Applied Mathematics, Computer Science, or a related field; or a Master's degree with 3-6 years of highly relevant experience.

Possess 0-3 years of post-PhD industry, startup, or postdoctoral experience, or 3-6 years total experience in computational science or engineering.

Hands-on experience implementing numerical methods for PDEs, such as FEM, FVM, FDM, particle or mesh-free methods, in research or production environments.

Experience with at least one major scientific computing or ML framework, for example JAX, PyTorch, or TensorFlow, and at least one GPU or performance-oriented technology such as CUDA.

Demonstrated success in speeding up simulations or building surrogate models for physics problems, with quantitative before/after results.

An "AI-first" workflow where LLMs are used to generate, refactor, and test code, enabling more time on modeling and physics.

U.S. citizenship, lawful permanent residency, or eligibility for required U.S. export authorizations to comply with ITAR and related regulations.

Strong foundational knowledge of linear algebra, calculus, and numerical analysis as applied to computational methods.

Ability to write correct, efficient code under tight deadlines while maintaining readability and thorough testing.

Experience with version control, continuous integration, and containerized workflows in a collaborative development environment.

Clear capacity to communicate complex technical concepts verbally and in writing to diverse audiences.

Willingness to iterate on feedback and refine implementations based on rigorous validation and team discussion.

Commitment to maintaining high standards of numerical integrity, reproducibility, and performance tracking.

Practical notes

This role requires U.S. citizenship or equivalent authorization due to export control requirements.

The position is based in Los Angeles, CA; Washington D.C.; or Seattle, WA, with details confirmed during the application process.

Typical interview steps

Hiring for engineering roles usually starts with a recruiter screen, followed by one or two technical rounds. Candidates often solve a coding problem, discuss past projects, and answer system design questions. Some loops include a take-home task. Final rounds typically cover team fit and give candidates a chance to ask questions. Interviewers look for how you break down an unfamiliar problem, not just whether you reach the answer. Practicing a few problems aloud and reviewing your own past projects are the best preparation.

Good to know

Open-source contributions and performance-oriented side projects are valued.

Questions to ask

Worth asking in any interview: how the team measures success, who the role works with daily, what the onboarding looks like, and what the company is trying to achieve this year. Asking what past hires did well is a strong final question. Keep the list short and pick the questions that matter most to you.

Career growth

Engineering careers usually progress from individual contributor to senior, staff, and principal levels. Some engineers move into management and lead teams of five to twenty people. Others stay on the technical track. Growth follows demonstrated impact, not tenure alone. A typical engineering ladder has clear levels with defined expectations for scope, quality, and mentorship. Moving up usually requires owning outcomes end to end rather than completing assigned tickets.