Staff Engineer, Machine Learning Systems & Reliability - Moveworks

We are building AI-enabled product capabilities that improve through data, feedback, and real-world use. We need the production systems that make those capabilities dependable: repeatable delivery, measurable quality, controlled learning loops, and reliable operation at scale. We’re looking for a hands-on Staff Engineer who can move machine-learning models, agentic workflows, and self-learning approaches from promising prototypes into secure, observable, continuously deployable production systems. This role sits at the intersection of ML systems, platform engineering, and site reliability engineering. You will partner with ML, data, product, and infrastructure teams to create a paved path from experimentation to production—and take ownership of how those systems perform and evolve once deployed. What you’ll do Design and build the production path for the complete ML lifecycle: data and feature preparation, training, experiment tracking, evaluation, artifact and model management, serving, monitoring, feedback collection, and retraining. Build continuous-delivery workflows for models, prompts, agent workflows, data dependencies, and supporting services. Establish automated quality, safety, performance, and compatibility checks. Implement safe rollout patterns such as shadow traffic, canaries, progressive delivery, feature flags, versioned artifacts, automated rollback, and operational kill switches. Turn self-learning approaches into controlled production feedback loops. Build systems for collecting outcomes, validating feedback, maintaining lineage, triggering model refreshes, comparing candidates, and promoting changes under explicit guardrails. Define and operate SLIs, SLOs, alerts, and error budgets across infrastructure, data pipelines, inference services, model quality, and product behavior. Connect model analytics and product telemetry with traditional operational signals so teams can understand whether a problem originates in infrastructure, data, model behavior, or the surrounding product. Improve the scalability, availability, latency, and cost efficiency of distributed training, inference, and data-processing workloads. Own capacity planning and resource optimization, including GPU resources where applicable. Participate in production ownership across the service lifecycle: architecture reviews, deployment, on-call, incident response, blameless postmortems, and systemic remediation. Build self-service platforms and automation that reduce operational toil and shorten the time required for ML engineers and data scientists to reach production. Apply LLMs or agentic automation to evaluation, troubleshooting, and operational workflows where they produce reliable, measurable improvements. Establish practical standards for cloud infrastructure, Kubernetes, infrastructure as code, observability, security, and compliance. Provide technical leadership across ML, data, product, and platform teams, mentoring engineers and influencing architecture without relying on formal authority. To be successful in this role you have: A track record of Staff-level technical ownership, typically gained through 7+ years of experience in software engineering, platform engineering, SRE, production engineering, or ML infrastructure. Strong software-engineering skills in Python and at least one production systems language such as Go, Java, C++, or Rust. Experience designing, operating, and troubleshooting distributed production systems, including failure analysis, capacity planning, and performance optimization. Hands-on experience with cloud infrastructure, containers and Kubernetes, infrastructure as code, CI/CD, and modern observability. Practical understanding of the ML lifecycle—including training, evaluation, model deployment, serving, monitoring, versioning, and retraining—and the ability to collaborate effectively with applied ML engineers or researchers. Experience distinguishing service-health problems from data-quality or model-quality p