Senior Data Labeler

Responsibilities: Create high-quality data that helps Runsphere evaluate model behavior and train better AI security controls.

• Label prompts, outputs, tool calls, traces, and outcomes against clear safety and quality criteria.
• Review difficult or ambiguous examples and document the reasoning behind each decision.
• Help develop taxonomies, guidelines, rubrics, and quality checks for new datasets.
• Identify recurring failure modes and surface them to research and engineering.
• Calibrate with other labelers, measure agreement, and improve consistency over time.
• Work carefully with sensitive or adversarial content within defined safety procedures.

Ideal candidate: You are precise, consistent, and comfortable making careful judgments about language and behavior. You can follow and improve detailed guidelines, explain decisions in clear English, and spot subtle patterns across many examples. You are reliable with repetitive work but curious about why systems fail. Experience with data labeling, moderation, QA, linguistics, research, AI, or security is a plus.