Overview
Our name is inspired by Theodore Roosevelt's 'Citizenship in a Republic' speech, honoring the '(hu)man in the Arena'. At Arena, we believe in fully committing ourselves to audacious, worthy causes, even in the face of potential failure.
We are a dedicated team of scientists and innovators reshaping the future of hardware engineering. If you are passionate about tackling real-world problems using fully autonomous AI, we invite you to join us in the Arena.
What we do:
In an increasingly complex physical world, the next generation of technology cannot rely on outdated tools. At Arena, we are crafting the world’s first AI industrial engineer to tackle the most intricate hardware and manufacturing challenges. Our flagship product, Atlas, leverages an in-depth understanding of physical systems and utilizes advanced knowledge in key physics domains. With the capacity to analyze and interpret multimodal industrial data, Atlas is adept at testing, debugging, optimizing, and repairing real-world products and systems. Trusted by leading industrial companies like AMD and Bausch & Lomb, Arena is expanding rapidly into sectors such as defense, automotive, and aerospace.
About the role:
As the Lead Research Scientist in Electromagnetism, you will spearhead the architecture, training, and validation of our electromagnetic foundation model, successfully combining numerically solved physics equations, neural operators, and high-quality simulations.
How you will contribute:
- Design innovative multiscale operator architectures with integrated physical constraints, including FNO, AFNO, Deformation-FNO, and U-FNO.
- Implement techniques for causality, passivity enforcement, and uncertainty calibration.
- Proprietary Training Data Development:
- Define coverage targets for data collection.
- Collaborate with Arena's Platform Engineering team to manage solver farms for synthetic data generation.
- Work alongside Arena Electrical Engineers to outline requirements for training data through hardware-in-the-loop testing.
- Create sim-to-real calibration methodologies utilizing advanced measurement techniques.
- Evaluation & Validation:
- Construct an automated suite for accuracy and constraint evaluation.
- Oversee releases and reporting procedures.
- Optimization & Performance:
- Analyze inference latency and throughput for performance improvements.
- Leadership & Documentation:
- Write design documents, experimental findings, and technical reports.
You have:
- A PhD or equivalent research experience in Electrical Engineering, Applied Physics, Computer Science, or Applied Mathematics.
- Demonstrated expertise in developing and training large foundation models.
- Experience with FNO and/or Neural Operators.
- A solid understanding of electromagnetic solvers (e.g., FEM, FDTD, MoM).
- Proficiency in Python, PyTorch/JAX, and C/C++ for bindings.
- A proven track record in physics-constrained machine learning or scientific simulation deployment.
- Experience in passivity/causality enforcement within learned models (preferred).
- Familiarity with power distribution networks, high-speed channel design, and EMI compliance testing (preferred).
- Knowledge of high-performance computing principles and familiarity with cloud service provider ecosystems (AWS, etc.) (preferred).
Benefits:
- 100% of the monthly premium for Aetna medical insurance, including vision and dental coverage.
- 401(k) Retirement Plan.
- Unlimited Paid Time Off.
- Daily lunch from local restaurants via Sharebite.
- Relocation assistance for positions in New York or San Francisco.
Salary: The base salary for this role ranges from $250,000 to $350,000 per year, depending on knowledge, skills, and experience. In addition to the base salary, we offer competitive equity and benefits packages.