Senior Software Engineer, Simulation

Atoms•San Francisco, CA

1d•$195,000 - $246,000•Onsite

About The Position

Atoms is seeking a Senior Software Engineer to build the log replay and scenario execution systems at the core of our simulation platform. In this role, you will take a recorded log from storage, to a replayable test case, to a scored result, and own the execution, determinism, and reporting in between. You will do this across on-road and off-highway programs, and you will own the tooling other engineers use to run scenario suites.

Requirements

Bachelor's degree in Computer Science, Robotics, Electrical Engineering, or a related field with 5+ years of software engineering experience, or a master's with 3+ years required.
Proficiency in C++ and Python, including the ability to read, navigate, and debug large codebases.
Experience building and operating data or test infrastructure at scale, including pipelines, batch execution, storage and indexing.
Experience with simulation, log replay, or large-scale automated testing for robotics or autonomous vehicles.
Working experience with a robotics middleware and its logging and replay tooling.
Experience handling autonomous driving or robotics sensor data across multiple modalities and timestamps, including camera, LiDAR, and radar.
Practical understanding of determinism and reproducibility, including clock control, seed management, and configuration pinning.
Experience integrating test workloads into continuous integration.
Working fluency with modern AI coding and agentic development tools.
Clear written communication covering design notes and test results.
Willingness to travel occasionally to deployment and test sites.
Ability to work onsite five days a week in San Francisco.

Responsibilities

Log replay: Build the pipeline that converts recorded logs into replayable test cases, including ingest, indexing, retrieval, and replay of a specific run on demand. Implement open-loop scoring against ground truth and closed-loop replay where the simulated machine is allowed to diverge.
Scenario execution: Build the execution layer that runs a scenario identically every time, including clock control, seed and configuration management, and results traceable to a software version.
Continuous integration: Integrate simulation into CI. Maintain the commit, nightly, and regression workflows, and keep failures attributable.
Metrics and reporting: Implement the pass criteria and metrics a release is judged against, and the reporting that reduces a batch of failing runs to an actionable list.
Scale and cost: Run large batches efficiently, track the cost of a simulation run, and reduce it as volume grows.
Tooling: Build the interfaces, visualizations, and command-line tools that make the platform usable by engineers who did not build it.
Cross-functional work: Work with autonomy and robotics on failure definitions, with bench and hardware-in-the-loop engineering on the path from simulation to bench to vehicle.