Find your next AI training role.

Annotate and review data used to train AI models, following a detailed labeling guide to keep output consistent across a large dataset. This is a 100% remote role sourced and manag...

Rate, rank, and review model-generated output against a written rubric, helping partners fine-tune their models toward what is actually helpful and accurate. This is a 100% remote...

Rate, rank, and review model-generated output against a written rubric, helping partners fine-tune their models toward what is actually helpful and accurate. This is a 100% remote...

Rate, rank, and review model-generated output against a written rubric, helping partners fine-tune their models toward what is actually helpful and accurate. This is a 100% remote...

Rate, rank, and review model-generated output against a written rubric, helping partners fine-tune their models toward what is actually helpful and accurate. This is a 100% remote...

Rate, rank, and review model-generated output against a written rubric, helping partners fine-tune their models toward what is actually helpful and accurate. This is a 100% remote...

Rate, rank, and review model-generated output against a written rubric, helping partners fine-tune their models toward what is actually helpful and accurate. This is a 100% remote...

Rate, rank, and review model-generated output against a written rubric, helping partners fine-tune their models toward what is actually helpful and accurate. This is a 100% remote...

Rate, rank, and review model-generated output against a written rubric, helping partners fine-tune their models toward what is actually helpful and accurate. This is a 100% remote...

Rate, rank, and review model-generated output against a written rubric, helping partners fine-tune their models toward what is actually helpful and accurate. This is a 100% remote...

Rate, rank, and review model-generated output against a written rubric, helping partners fine-tune their models toward what is actually helpful and accurate. This is a 100% remote...

Probe a production AI system for unsafe, biased, or policy-violating output using structured adversarial techniques, then document findings clearly for the partner's safety team. T...