JOB-1604
A fast-growing company that supplies training data to frontier AI labs is hiring its first dedicated quality specialist. At present team leads do quality checks alongside their regular duties; this hire creates a separate audit layer that inspects each dataset prior to shipment and answers to operations leadership.
Responsibilities
- Hold the last quality gate before anything is handed over, inspecting datasets of code, reasoning and RL trajectories for accuracy, structure and conformity to spec
- Build the quality function from zero, including how reviews run and the rubrics and scoring standards used
- Collaborate with team leads and delivery groups to uncover systemic problems and route corrections into contributor training
- Team up with platform engineers to automate quality workflows and tools so errors do not recur
- Track quality metrics and tell operations leadership how often errors occur, why they happen and where things are trending
Ideal background
- Two to ten years as a software engineer, data scientist or ML engineer, with hands-on quality assurance on technical or human-generated data as a core, full-time duty
- Ability to review, debug and judge code and technical data
- A keen sense of what excellent data is and the independence to work unsupervised
- Plus: experience at an AI data company, RLHF or annotation quality work, rubric design, automated check scripts in Python, or a statistics background
On-site in the San Francisco Bay Area. Visa transfers considered.
