Human Data Architect

Highlights

• Design data systems governing how human data trains frontier AI models.
• Build RL environments and rigorous evaluations that go beyond standard benchmarks.
• Partner with engineering to improve pipelines without sacrificing data quality.

Ideal candidate

Systems thinker with background in cognitive science, HCI or crowdsourcing who cares how human data shapes model behavior and loves designing rigorous collection methods at scale.

Requirements

• Experience designing human data collection at scale, e.g. in computational cognitive science, HCI or crowdsourcing research
• Familiarity with ML post-training paradigms such as SFT and RL and how data quality maps to model outcomes
• Strong technical skills with ability to work across data pipelines, not just analyze endpoints
• Systems-level thinking about how task design choices propagate into training data
• Excellent communication skills explaining measurement to engineers and constraints to researchers