Software Engineering Resident

San Francisco
ML Team /
Extropic is looking for a Software Engineer with an R&D focus to join our residency program. Our hardware massively accelerates certain kinds of probabilistic inference, and residents will help pioneer the software to design these thermodynamic models and the engineering of training them.


    • Collaborate with senior researchers and engineers to develop and improve our software stack.
    • Implement new algorithms designed to simulate our thermodynamic hardware.
    • Conduct tests and experiments to evaluate the effectiveness of our thermodynamic models and other software solutions.
    • Maintain comprehensive documentation of our software stack and provide troubleshooting and internal support to researchers.

Required Qualifications

    • Strong experience with JAX is mandatory
    • Experience with debugging, maintaining, deploying python packages
    • Experience with scientific Python
    • Experience with version control tooling (e.g. git) and CI/CD tooling (e.g. GitHub actions)
    • Experience with writing and designing software documentation, and deploying it using standard tools (e.g. sphinx and readthedocs)
    • Knowledge of probability and linear algebra
    • Experience with infrastructure for machine learning experimentation and training in parallel/distributed environments (e.g. Slurm, Weights & Biases, etc.)

Preferred Qualifications

    • History of open source contributions to machine learning or scientific frameworks
    • Experience with numerical methods (e.g. differential equation solving, monte carlo integration, automatic differentiation etc.)
    • Strong grasp of computational Bayesian methods, including MCMC sampling methods, variational inference, and energy-based models
    • Publications in ML or computational science venues

    • Few people will have all of these skills, so anyone with partial overlap is encouraged to apply!
$75,000 - $150,000 a year
Salary will vary with experience
Extropic is an equal opportunity employer