Mid-Level Research Engineer
Core
Build and improve scalable systems that automate quality control for reinforcement-learning training data.
Role type
Mid-level research engineer (RL data quality)
Builds
Automated QA/QC systems, validation pipelines, and internal tools for RL training data
Domain
Artificial Intelligence / Reinforcement Learning / Data Engineering
Deliverable
production ML models
Required skills
Python, Docker, Linux, statistics, data-validation, automated QA/QC systems, benchmarks and evaluations for RL data, sampling strategies, rule-based validation, model-assisted validation
Preferred skills
Experience with realistic tasks, reliable rubrics, and useful training trajectories
Technologies
Python, Docker, Linux
Responsibilities
Design experiments and metrics to evaluate agent outputs and training data; Investigate quality issues with data vendors and diagnose agent failures; Develop sampling strategies, audits, and validation pipelines; Integrate quality checks into internal tools and vendor workflows
Seniority
Mid-level, hands-on IC
