ML Ops Engineer at Island

The role

Island makes an enterprise workspace that combines access, security and AI productivity, and machine learning models sit behind its AI security features, some trained in-house and others open-weight community models. As an ML Ops Engineer in Tel Aviv, you are responsible for getting models from research into production through packaging, release and serving, and for keeping them fast, reliable and affordable. You work closely with researchers and backend engineers.

What you will do

• Build and run the serving infrastructure for production models, on AWS SageMaker and other clouds.
• Run model releases, including versioning, packaging, CI/CD, careful rollouts and rollbacks.
• Host open-weight models and LLMs on CPUs and GPUs.
• With the Research team, create the pipelines for training jobs, datasets and evaluation.
• Monitor latency, errors, quality and drift in production, and set up alerts.
• Control compute costs by choosing suitable instance sizes and tracking GPU spending.

What you need

• 4 or more years of backend, platform or DevOps engineering, with real ownership of production systems.
• Experience taking ML models into production. Training models is not required, but you should know how models get built, served and evaluated.
• Strong cloud skills, especially AWS (IAM, S3, networking).
• Infrastructure as code with Terraform or similar.
• CI/CD pipelines with GitHub Actions or similar, and Docker.
• Daily use of AI coding tools.
• A team-minded approach, a sense of ownership and the ability to learn quickly.

Nice to have

• AWS SageMaker endpoints, training jobs or pipelines.
• ML services on other clouds, such as Azure ML or Vertex AI.
• Serving LLMs with tools such as vLLM, TGI or Triton, and running GPU instances.
• Model optimization with ONNX or quantization.
• Kubernetes.
• Workflow tools such as MLflow, SageMaker Pipelines or Airflow.

About Island

Island says millions of people at large enterprises use its product. Its investors include Coatue Management, Insight Partners, Sequoia Capital and Cyberstarts.