Staff+ Site Reliability Engineer, Safeguards ML Infra
Looking for a machine learning engineer job in Seattle? Choose from 25 open roles at 12 employers. Explore real salary data and skills breakdown across remote, hybrid and on-site roles.
Free. No credit card needed.
The work spans a broad slice of applied machine learning. Several roles centre on building and maintaining production systems: deploying models, managing inference pipelines, handling on-call operations, and automating operational processes, as seen in the Anthropic safeguards infrastructure role and the Amazon data centre analytics role. Others sit closer to research and modelling: designing computer vision systems for object detection and video analytics, as in the Metropolis computer vision role, or architecting on-device, edge-deployed models for safety-critical hardware, as in the Axon principal engineer role. Platform work also features, with roles focused on model integration, inference services, and storage systems supporting large consumer products, such as the Adobe platform engineering role. Advertising and monetisation appear repeatedly in the sample titles, suggesting that recommender and ranking systems are a meaningful thread across these postings. The Metropolis senior manager role adds a people-leadership dimension involving MLOps infrastructure and agentic AI deployment.
These counts reflect mentions in role descriptions across the postings on this page, not formal requirements, and the sample is small enough that any single posting can shift a count noticeably. LLMs and Python appear in most postings. PyTorch, evaluation methodology, SQL, and A/B testing appear in roughly half. A PhD and prior publications are mentioned in a meaningful share of postings, though whether they are required or preferred varies by role and is worth checking in each listing.
The clear majority of postings on this page are listed as fully on-site. Only a small minority state hybrid arrangements, including the Axon principal engineer and the Metropolis computer vision and manager roles, which specify four days weekly on-site. A very small number are listed as remote. No postings have an unknown work mode. Candidates looking for flexibility should check each listing carefully, as on-site expectations vary in how strictly they are stated.
Jobs checked 1 day ago. · Guide reviewed 9 September 2026.
Search by role, company, or anything a posting mentions.