DevOps Engineer

DevOps Engineer | AI SaaS

New York City Hybrid, 2 day a week onsite $120,000-$160,000 base

How do you keep an AI feature responsive when demand spikes, inference gets slower, and some jobs take minutes to finish? You'll build the infrastructure that handles those trade-offs for a customer-facing SaaS product.

You'll work directly with software and ML engineers to get model-backed features into production. The team needs someone who can trace a slow request through an API, a queue, and a model endpoint, then decide whether to change how the workload scales, how jobs are scheduled, or how the application handles failures.

What you'll build

Infrastructure for different workloads. Run customer-facing APIs, inference services, and batch workers on Google Cloud Platform and GKE, with scaling rules that reflect how each workload behaves.

A repeatable path to production. Give engineers a way to deploy and roll back application and model-serving changes independently, with clear version tracking.

Visibility into latency and cost. Use OpenTelemetry and workload metrics to understand where requests slow down and how much different workloads cost to run.

Tools for the engineers shipping AI features. Build Python or Go tooling that helps teams provision environments, deploy services, and troubleshoot workloads themselves.

What you'll bring

You've operated Kubernetes in production, written infrastructure as code, and debugged systems with APIs, queues, and workers. You can write automation and use production data to make decisions about latency, throughput, and cloud spend.

Experience with model serving, GPUs, or high-volume job processing is welcome. A background in scaling SaaS infrastructure is also relevant.

Apply with your rsum. We'd be interested in a performance or scaling problem you've worked through and the trade-offs you made.

IT - DevOps Engineer - New York CIty

Oscar Associates Limited (US) is acting as an Employment Agency in relation to this vacancy.