Scientific Computing Engineer, Numerical Solvers

The Role

We're looking for a Scientific Computing Engineer to own and develop significant parts of our in-house numerical solver, from mathematical formulation to tested, production-quality code behind a stable interface: numerical methods, iterative solvers, validation and performance optimization amongst others.

Working closely with our technical lead, you'll shape how the solver is formulated, implemented and run on GPU and CPU hardware. You'll work with the algorithm team, who use the solver, and with the software team, who integrate and deploy it in our product. You own your parts of the solver end to end, and we expect you to challenge existing designs when you see a better approach.

Key Responsibilities

• Design and implement the solver's numerical methods, and improve their accuracy and speed
• Build and maintain the testing, validation and benchmarking that keep the solver's results accurate and reliable
• Optimize performance on GPU and CPU hardware, together with the technical lead
• Work with the algorithm team so that the solver meets the needs of our measurement system, and with the software team, who integrate and deploy it in the product
• Contribute to technical decisions, code reviews and documentation, and challenge existing designs when you see a better approach
• Take on adjacent responsibilities as the team grows, such as contributing to the longer-term solver roadmap

Is this you?

• PhD in physics, applied mathematics or computational science, or equivalent industry experience, and 3 to 5 years of experience developing numerical software in research or industry
• Strong background in numerical methods and numerical linear algebra, for example integral equations, iterative (Krylov) solvers and preconditioning
• Understanding of how code maps to hardware (memory, data layout, performance), typically gained through a lower-level language such as C, C++, Rust or Fortran
• Fluent English, and able to explain your technical choices clearly to colleagues with different backgrounds
• Active use of AI tooling in your daily work (coding, testing, benchmarking, documentation) to boost efficiency

Nice to have

• A background in computational physics with similar mathematics, for example electromagnetism, optics, wave propagation, quantum mechanics, condensed matter, many-body physics or plasma simulation
• Experience with Python and/or Julia
• Experience with GPU programming or high-performance computing
• Software engineering practice in a team: version control, testing, code review, and code that others can maintain
• Experience with numerical optimization or adjoint methods

What We Offer

• A role at the heart of Invisix's numerical software, with real technical ownership from day one
• Competitive salary and benefits, including an Employee Share Ownership Plan
• A learning and development budget to support your growth
• Flexible working arrangements, and room to grow into adjacent responsibilities as the team scales