SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 170,000 - 210,000 / annual
Hungryroot is an AI-powered food and wellness company that acts as a personal assistant for healthy living. The company recommends and delivers personalized groceries, recipes, and supplements to customers based on their goals, lifestyle, and budget. Hungryroot operates as a distributed, remote-first team across 28+ U.S. states with headquarters in New York City.
The Senior Machine Learning Operations Engineer will join the Data Science team, which owns production systems powering grocery recommendations and box personalization. The platform combines Python services, FastAPI APIs on AWS, Spark pipelines on Databricks, and ML models feeding a real-time decisioning engine.
Key responsibilities include designing and operating scalable backend services and data pipelines; improving reliability, performance, and observability of production ML systems; owning the model-to-production lifecycle including versioning, safe rollout/rollback, and monitoring for data quality and drift; building clean interfaces for ML model integration with experimentation and feature-flag tooling; strengthening engineering foundations through testing, type checking, CI/CD, and infrastructure as code; profiling and optimizing data-heavy services; and collaborating with data scientists, operations researchers, and product engineers.
Required qualifications: 5+ years in MLOps, ML engineering, or DevOps focused on production ML infrastructure; strong Python and SQL with Bash automation; experience designing and operating backend services and APIs with attention to reliability and scalability; hands-on Databricks and Spark experience; CI/CD experience for ML/data systems; solid AWS fundamentals including IAM, networking, and containerized workloads; and production observability experience including ML-specific monitoring.
Nice-to-have skills include familiarity with recommendation and personalization systems, optimization solvers, experimentation platforms, feature stores, low-latency model serving patterns, and cost/performance optimization of data workloads.
