Apply trust and safety judgment remotely as a Data Annotation Specialist, evaluating AI outputs and creating accurate safety annotations for Cohere.

Cohere seeks Data Annotation Specialists with experience in content moderation, trust and safety, AI data annotation, or related analytical work. You will label, rank, audit, and refine human-generated and model-generated text. The work improves safety, quality, and policy alignment across complex scenarios and user intents.

Key Responsibilities:

- Evaluate model outputs against detailed safety guidelines and rubrics
- Label, rank, audit, and refine generated text
- Create prompts, user scenarios, and adversarial safety examples
- Identify annotation inconsistencies and recurring model failures




- Complete tasks independently with careful attention to detail

Requirements:

- One or more years in moderation, trust and safety, or annotation
- Experience applying guidelines to complex, sensitive content
- Comfort evaluating sexual, violent, or psychologically disturbing material
- Excellent written English and clear evaluation explanations
- Strong contextual judgment, accuracy, and personal bias awareness

Use contextual judgment, written English, and safety evaluation skills to improve Cohere’s models through careful data annotation.

📌 Trust and Safety Data Annotator (Vancouver)
🏢 Cohere
📍 Vancouver