Machine Learning Engineer (Staff) (San Francisco)

Staff Machine Learning EngineerAbout Sprinter HealthAt Sprinter Health, our mission is reimagining how people access care by bringing it directly to their homes. Nearly 30% of patients in the U.S. skip preventive or chronic care simply because they can’t get to a doctor’s office. For many, the ER becomes their first touchpoint with the healthcare system, driving over $300B in avoidable costs every year.By using the same technologies that power leading marketplace and last-mile platforms, we deliver care where people are, especially those who need it most. So far, we’ve supported more than 2 million patients across 22 states, completed 130,000+ in-home visits, and maintained a 92 NPS. Our team of clinicians, technologists, and operators has raised over $125M from investors like a16z, General Catalyst, GV, and Accel and enjoys multi-year runway.About the RoleWe’re looking for a Staff Machine Learning Engineer to be Sprinter’s first dedicated ML engineering hire and build the production systems that train, deploy, monitor, retrain, and serve machine learning models across the company.This is a founding, first-of-function role. You will define the blueprint for how ML moves from prototype to production at Sprinter, including our training and inference pipelines, serving patterns, feature workflows, monitoring, validation, retraining, and model governance practices.You’ll work closely with engineering, data, product, operations, and applied science teams to turn models into reliable systems the company can depend on. That includes serving predictions through APIs and batch jobs, building clean interfaces between data and product systems, and implementing the observability needed to catch drift, data quality issues, latency problems, cost regressions, and silent model degradation before they impact patients or operations.Just as importantly, you’ll make the foundational calls that every future model and ML engineer will build on: build versus buy, serving architecture, feature paradigms, deployment standards, monitoring expectations, and the guardrails that allow us to move quickly without creating fragile systems.This role is ideal for a staff-level, hands-on engineer who thinks in systems, has built ML infrastructure from the ground up, and knows how to right-size solutions for a rapidly growing startup. You should be someone who empowers the teams around you, accelerates time to deployment, and knows what a model needs to be truly production-ready.As the function grows, you will have the opportunity to shape the team, define the technical bar, and help build the ML engineering foundation for Sprinter.Office LocationWe are a hybrid company based in the Bay Area with offices in both San Francisco and Menlo Park. We operate on a hybrid schedule, working from the office Monday through Thursday, with Fridays designated as work-from-anywhere days.We care deeply about work-life balance and are happy to provide flexibility when life happens. We ask that employees be in the office Monday through Thursday to collaborate with their teams while maintaining flexibility where it matters most.Lunch is provided every day, and the entire team takes an hour to eat together. It’s one of the ways we stay connected outside of meetings. You’ll usually find us playing a board game before getting back to work.What you will doBuild and lead Sprinter’s ML engineering function as the company’s first dedicated ML engineering hireDefine Sprinter’s ML platform and deployment paradigm across training, serving, features, monitoring, retraining, and governanceMake foundational build-versus-buy, architecture, tooling, and platform decisions that future models and engineers will build onDesign and build production training and inference pipelines that are reliable, observable, and maintainablePackage models for deployment and serve predictions through APIs, batch jobs, or other production workflowsBuild clean interfaces between data systems, models, and product systems so ML can be consumed safely and reliablyMaintain feature pipelines and ensure features remain fresh, correct, and consistent between training and servingImplement monitoring for model performance, drift, data quality, latency, cost, reliability, and production behaviorPrevent training-serving skew, silent degradation, and model regressions before they become production issuesAutomate retraining, validation, deployment, rollback, and other production ML workflows where appropriateEstablish reproducibility, versioning, model governance, and operational readiness practices as company defaultsPartner with engineering, data platform, product, operations, and applied science teams to productionize models and improve handoffsWrite design docs, define technical standards, and bring the broader engineering organization along on key ML infrastructure decisionsSet the technical bar for ML engineering by helping interview, mentor, and eventually hire engineers who followWhat you have doneSpent 8+ years building production software, data systems, ML systems, platform infrastructure, or related technical systemsBuilt and owned ML systems in production across training, serving, features, monitoring, and deploymentTaken models from prototype or research stage into reliable, production-grade systemsBuilt or meaningfully scaled ML infrastructure, MLOps platforms, model-serving systems, feature pipelines, or related infrastructureDesigned systems that other engineers, data scientists, analysts, or product teams rely onMade architectural decisions around ML platform design, serving patterns, feature infrastructure, build versus buy, and operational standardsWorked with cloud infrastructure, containers, CI/CD, orchestration, data pipelines, and production deployment workflowsBuilt monitoring, observability, validation, or alerting for ML systems, data systems, or high-reliability production servicesCreated reproducible workflows across data, features, models, training runs, deployments, or experimentsPartnered closely with data science, applied science, data platform, product, operations, or backend engineering teamsOperated in ambiguous environments where there was no existing playbook and technical decisions had a long half-lifeBalanced speed, simplicity, reliability, privacy, and long-term maintainability in production systemsWhat gives you an edgeYou’ve been an early ML engineer, founding ML engineer, or first ML infrastructure hire at a startupYou’ve built ML infrastructure in a high-growth or operationally complex environmentYou have depth in large-scale model serving, feature infrastructure, LLM infrastructure, or real-time inference systemsYou have a background in backend engineering, data engineering, MLOps, platform engineering, or infrastructure engineeringYou have experience with feature stores, feature pipelines, or production data systems at scaleYou’ve helped interview, hire, mentor, or set the technical bar for ML engineers, platform engineers, or data engineersYou’ve worked with healthcare data, PHI, HIPAA-aware systems, or other sensitive data environmentsYou have experience with security, privacy, governance, or compliance considerations for production ML systemsWhat makes you successfulYou decide what the pattern should be and bring the rest of the organization alongYou reach for the simplest system that works, adding complexity only when the value justifies itYou know what it takes to make a model production-ready and can communicate those requirements clearlyYou are an accelerator for applied science, data, product, and engineering teams, not a gatekeeperYou build interfaces that make models easy to consume and hard to misuseYou prevent silent degradation before it becomes an incidentYou create standards that help future engineers move fasterYou raise the technical bar for everyone who joins the function after youDay to DayIn this role, you might spend your time:Deciding what Sprinter’s serving and feature paradigms should be and writing the design docs behind those decisionsHardening a training pipeline or batch-inference workflowProductionizing a model handed off from another teamDebugging a model-serving issue or production data quality problemReviewing feature freshness, model performance, drift, latency, or costBuilding validation and rollback workflows for model deploymentsPartnering with product and operations teams to understand how model behavior impacts real-world workflowsInterviewing a candidate, mentoring an engineer, or setting a new technical standard for the ML engineering functionThe Interview ProcessWe aim to complete the interview process within 2–3 weeks. It will usually consist of:Recruiter Screen: Background fit, motivation, and compensation alignmentHiring Manager Interview: Technical experience, first-of-function fit, and ML infrastructure depthHands-on Technical Assessment: Practical ML engineering, production systems, and implementation abilityOnsite Interview: Systems design, technical case study, behavioral interview, and lunch with the teamReferences: Validation of performance, judgment, and working styleWhat we offerMeaningful pre-IPO equityMedical, dental, and vision plans 100% paid for you and your dependentsFlexible PTO + 10 paid holidays per year401(k) with match16-week parental leave policy for birthing parent, 8 weeks for all other parentsHSA + FSA contributionsLife insurance, plus short and long-term disability coverageFree daily lunch in-officeAnnual learning stipendRelocation assistanceEqual Opportunity StatementSprinter Health is an equal opportunity employer. We value diversity at our company. We do not discriminate on the basis of race, religion, color, national origin, gender, sexual orientation, age, marital status, veteran status, disability status, or other protected classes.Beware of recruitment fraud and scams that involve fictitious job descriptions followed by false job offers.If you are applying for a job, you can confirm the