Staff Infrastructure Engineer, Robotic Systems

Atomic Machines•Santa Clara, CA

1d•Onsite

About The Position

Atomic Machines is building the Matter Compiler™, an AI-native, all-digital manufacturing system that builds machines directly from code. The company’s first product is PrimeSwitch™, an electromechanical power relay. Atomic Machines is based in the San Francisco Bay Area, with offices in Emeryville and Santa Clara. The Staff Infrastructure Engineer will own end-to-end infrastructure for the Matter Compiler fleet, including the multi-package monorepo build system, CI/CD pipeline, artifact store, shared cache, and deployment path to approximately 80 nodes today and thousands of devices across several metros by 2027. The role covers designing, building, operating, and improving this infrastructure; deciding what to automate; setting safeguards that limit the impact of bad builds; defining pipeline-health signals; and driving adoption of infrastructure across teams. The role requires decomposing ambiguous system failures, reasoning across abstraction layers, and owning problems that cross team and organizational boundaries. Hardware experience is not required. The position requires working onsite five days per week.

Requirements

Experience owning a build system for a repository with many packages and multiple build targets, including identifying problems and making changes.
Experience owning a CI/CD system end to end, including rollback, and an understanding of how deployments can be made safer over time.
Deep understanding of build reproducibility and caching, including local-versus-CI test discrepancies, cache-key design, and the possibility of incorrect artifacts despite a correct key.
Deep understanding of fleet deployment, including risks of mutable tags, deployment by digest, and safe production rejoining by offline nodes.
Experience designing concrete failure paths, including timeouts, safe resulting states, and information presented to the next engineer.
A record of diagnosing production incidents to root cause and implementing a mechanism, measurement, and fix.
A record of getting other engineers to adopt built infrastructure and an understanding of the lever used to drive adoption.
Candor about areas of weakness and ownership of outcomes, including unsuccessful ones.
Fluency with AI tools in daily engineering work, while taking ownership of decisions, code, and results.
Ability to work onsite five days per week.

Nice To Haves

Experience deploying software to distributed nodes that control physical equipment.
Experience with GitLab CI, Ansible, OIDC/STS-based credential flows, and artifact stores at scale.
Experience operating infrastructure through growth from tens to thousands of endpoints.

Responsibilities

Own the build system for a 40-package monorepo targeting three architectures, including hermetic builds, lockfiles, change detection, and a shared remote cache keyed on real inputs.
Own CI/CD end to end, including rollback; measure push-to-green time, queue depth, and fetch error rate, and improve those metrics.
Design and operate the fleet deployment path, including immutable artifacts, staged rollouts, version handshakes, drift detection, and an abort path independent of the deploy script.
Define production signals and thresholds for pipeline health and create runbooks that on-call engineers can use independently.
Convert runbook steps into pipeline checks and automated remediation where worthwhile.
Own build-runner security as a blast-radius problem, including short-lived credentials, scoped IAM, credential rotation, and secret scanning in CI.
Work with IT on GitLab, network, and identity boundaries; clarify boundaries rather than making assumptions.
Remove and measure manual toil, tracking what engineers did by hand before and after changes.
Drive adoption of infrastructure across teams that do not report to the role, making improved workflows the default.
Raise reliability standards without blocking delivery and encode lessons from incidents into systems.
Diagnose production incidents to root cause and implement mechanisms, measurements, and fixes.
Use AI tools in daily engineering work while owning accepted decisions, code, and results.