CareerPlanSign in

Staff Engineer, Test Automation (R5792)

San Diego, California💼 Full-time🗓 2026-09-26 → 2026-09-29

Core

Own technical strategy and architecture for automated testing, verification, and MLOps quality across autonomy software, backend services, APIs, and distributed hardware environments.

Role type

Staff Engineer, Test Automation & MLOps

Builds

Scalable test frameworks, automated ML validation pipelines, CI/CD workflows, and observability tools for autonomy systems.

Domain

Defense technology, autonomous systems, machine learning, distributed computing

Required skills

Python automation, MLOps, CI/CD, distributed systems testing, GPU workload testing, Kubernetes validation, hardware-software integration testing, asynchronous programming, observability, Linux development

Preferred skills

MLflow/Kubeflow, GPU orchestration (Run:ai, Kueue), NVIDIA CUDA/PyTorch, embedded/edge testing, fault-injection testing, infrastructure as code, Go/TypeScript, C/C++ bindings, aerospace/robotics domain knowledge

Technologies

Python, Kubernetes, MLflow, Kubeflow, Weights & Biases, SageMaker, Vertex AI, Run:ai, NVIDIA Nsight, PyTorch, Conan, Poetry, Nix

Responsibilities

Design scalable test frameworks for autonomy software and distributed hardware; lead functional, integration, regression, and performance testing across simulation and HIL/SIL environments; build automated ML validation pipelines for data quality and model robustness; establish CI/CD and continuous training workflows; develop scenario-based validation for edge cases and distribution shifts; create observability and failure-triage capabilities; build Python automation for test execution and developer productivity; develop test harnesses, simulators, and synthetic data capabilities; collaborate with engineering teams on verification strategies; develop AI-assisted engineering workflows using coding agents and LLMs.

Seniority

Staff, hands-on IC with strategic ownership

Sourced via lever · Listed on CareerPlan, which tracks 872,000+ jobs from 20+ sources.