Software Engineer, Workflow Systems

NVIDIA•Santa Clara, CA

2d•$152,000 - $287,500

About The Position

NVIDIA is building software to make chip-design workflows repeatable, inspectable, and recoverable. These workflows coordinate long-running tools, generated design files, shared compute, and dependencies across engineering teams. The role primarily develops workflow definitions, configuration models, and reusable tool interfaces, working with workflow owners, chip-design specialists, and runtime engineers. The team will deliver usable RTL-to-GDS workflows and then extract reusable capabilities from those implementations. Runtime engineers own the underlying execution platform; this position builds workflow software and integrations that use it.

Requirements

BS/MS in Computer Science, Computer Engineering, Electrical Engineering, or equivalent experience.
At least 4 years developing production software, or equivalent demonstrated scope.
Strong programming skills in Go, Python, C++, Java, Rust, or another suitable language, and ability to learn the team's languages.
Ability to demonstrate clear interfaces, state models, meaningful tests, and reasoned implementation tradeoffs.
Experience with workflows, build systems, job execution, release systems, scientific or ML infrastructure, controllers, or other software that manages dependent work and persistent results.
Concrete reasoning about retries, concurrent activity, incomplete outputs, process failure, and safe changes to systems already in use.
Ability to learn an unfamiliar domain, make implementation decisions independently, and carry a change through verification and delivery.

Nice To Haves

Owned worker or scheduler integrations, workflow execution, isolation, or recovery mechanisms under real failure conditions.
Built schemas, configuration layering, compilers, build graphs, or APIs that other engineers successfully extended.
Implemented reproducible runs, artifact manifests, lineage, or cache invalidation.
Modernized mature systems with compatibility tests, shadow comparisons, controlled adoption, and rollback.
Debugged long-running tools and shared compute using Linux processes, files, permissions, resource accounting, or equivalent runtime mechanisms.
Integrated EDA tools or worked with design, verification, HPC, or research teams.
Prior chip-design experience is preferred but not required; relevant experience from build systems, scientific computing, HPC, or research tools can transfer to the role.

Responsibilities

Build workflow definitions, dependency models, and configuration that engineers can understand, making final settings, their origins, and their effects explainable.
Implement workflow state and connect it to worker coordination, cancellation, retries, recovery, and resource controls.
Collaborate with runtime engineers on isolation and artifact publication to protect shared inputs and reject stale or incomplete results.
Integrate tools and schedulers through versioned interfaces with explicit inputs, outputs, resource needs, and completion checks.
Expose workflow and attempt state with diagnostics that identify the failing stage and the next useful action.
Improve existing Python, Go, Perl, Tcl, Make, and shell infrastructure through focused changes, compatibility tests, and staged rollout.
Partner with workflow owners to validate complete scenarios, reproduce failures, and support production adoption.
Use AI development tools effectively, review generated code, independently verify expected test results, and own the behavior shipped to users.