Research Engineer

Location: San Francisco Bay AreaWork model: On-site (some team members are remote, but this role is currently on-site)Industry: AI infrastructure / Reinforcement Learning (RL) training data & evaluationsCompensation: Competitive (range not provided) + benefits (medical/dental/vision coverage, meals, 401(k), commuter benefits, wellness perk)About the Company (our partner)Our partner is a fast-growing, venture-backed AI infrastructure company building the tooling and workflows that power reinforcement learning (RL) training data and evaluation for frontier AI agents. Their platform is used by advanced AI teams across large enterprises and high-growth startups, and they're scaling quickly to meet strong customer demand. The team is small, highly technical, and execution-focused, with a culture that values ownership, speed, and craftsmanship.The OpportunityOur partner is hiring a Research Engineer to help scale the quality assurance (QA) systems behind training data generated through their infrastructure. This role sits at the intersection of data quality, tooling, and applied ML operations: you'll build the standards, pipelines, and feedback loops that ensure datasets are reliable, consistent, and ready for training and evaluation.You'll work closely with internal stakeholders and external data suppliers to diagnose quality issues, improve workflows, and continuously fold QA learnings back into the platform. If you enjoy building systems that make high-quality data scalable—and want to do it in a high-ownership, fast-paced environment—this role is a strong fit.ResponsibilitiesDefine and enforce quality standards for training datasets used for RL training and evaluationBuild tooling and workflows to audit supplier-generated datasets, including sampling strategies, validation pipelines (rule-based and model-assisted), and feedback loopsEvaluate and implement human-in-the-loop review workflows where beneficial to improve quality and efficiencyPartner with external data suppliers to debug quality issues, provide actionable feedback, and improve their data generation processesIntegrate QA learnings into internal tools and supplier portals to reduce anomalies, inconsistencies, and edge cases over timeTrack QA outcomes and continuously improve processes, metrics, and documentationRequirementsProficiency with Python and experience working in Linux environmentsExperience with Docker and reproducible development/deployment workflowsExperience working with large-scale datasets (validation, transformation, or analysis)Strong problem-solving skills and evidence of rapid learning in technical environmentsAbility to operate independently and deliver results in an early-stage, fast-moving settingClear written and verbal communication skills (including collaborating across time zones)Nice to haveExperience building data validation pipelines and/or human-in-the-loop review systemsFamiliarity with common training-data failure modes and techniques to detect subtle inconsistenciesComfort designing QA metrics, experiments, and processes—not just executing predefined checksFamiliarity with modern AI tooling and LLM capabilities