Remote Research Scientist, Electromagnetic Inverse Design
Flexcompute
Job highlights
Summary
Join Flexcompute as a research scientist to drive the development and adoption of adjoint-based inverse design using Tidy3D electromagnetic solver, working on projects spanning algorithm design, software development, and practical application.
Requirements
- PhD or Master's degree in a relevant field (Photonics, Physics, Electrical Engineering, Machine Learning, etc.) with focus on inverse design, optimization, or machine learning
- Excellent programming skills, specifically in Python
- Good understanding of computational methods and numerical simulations
Responsibilities
- Support development of new features and performance improvements within Tidy3D’s differentiable simulation framework
- Assist in design and implementation of APIs to make inverse design simple and easy to use
- Apply inverse design to practical problems in industry and academic collaborations, craft documentation examples, and understand real world impact and use of our tool
Preferred Qualifications
- Experience with practical realization of inverse design, spanning from simulation to experiment
- Industry experience in roles involving inverse design
- Experience with developing numerical simulation code
- Machine learning experience, especially familiarity with automatic differentiation tools such as JAX or PyTorch
- Experience implementing adjoint method for calculating gradients through physical simulations
- Experience with Tidy3D as a user
Benefits
- Fully remote work arrangement
- Competitive compensation
Job description
Flexcompute is seeking a research scientist to be a key driver of the development and adoption of adjoint-based inverse design using the Tidy3D electromagnetic solver. The candidate will work on various projects spanning from algorithm design, software development, and targeting the practical application and adoption of our inverse design platform. This candidate will work in a tight knit team and will be instrumental in our push to make Tidy3D the go-to tool for inverse design in electromagnetics.
Key responsibilities include:
- Support development of new features and performance improvements within Tidy3D’s differentiable simulation framework, including new physics and new optimization algorithms.
- Assist in design and implementation of APIs to make inverse design simple and easy to use.
- Apply inverse design to practical problems in industry and academic collaborations, craft documentation examples, and understand real world impact and use of our tool.
Required:
PhD or Master’s degree in a relevant field (Photonics, Physics, Electrical Engineering, Machine Learning, etc.) with focus on inverse design, optimization, or machine learning.
Excellent programming skills, specifically in Python.
Good understanding of computational methods and numerical simulations.
Nice to have:
Experience with practical realization of inverse design, spanning from simulation to experiment.
Industry experience in roles involving inverse design.
Experience with developing numerical simulation code.
Machine learning experience, especially familiarity with automatic differentiation tools such as JAX or PyTorch.
Experience with open source and collaborative software development process, ideally in industry setting.
Experience implementing adjoint method for calculating gradients through physical simulations.
Experience with Tidy3D as a user.
Fully remote, with collaboration with a team based primarily in US and European time zones.
Applicants based outside of US considered.
Competitive compensation.