Work Type
Remote
Location
Global
Compensation
$20 - $50 USD per hour
About this opportunity
Overview
TELUS Digital is looking for experts to support reinforcement learning from human feedback (RLHF) work that helps align AI systems with human values. In this role, you would design evaluation strategies and grading rubrics that assess AI responses for logic, tone, and instruction-following across different cultural contexts. The work is remote and part of TELUS Digital’s global AI community.
What You’ll Do
- Create multi-dimensional grading rubrics for evaluating AI responses.
- Review outputs for tone, reasoning quality, and adherence to instructions.
- Help audit large-scale annotation and quality-assurance workflows.
- Investigate hallucinations and logical fallacies in model outputs.
- Develop prompt-engineering strategies to reduce recurring issues.
- Work with project stakeholders to define the ideal response profile for specific industries or user personas.
Requirements
- Minimum bachelor’s degree in a technical field such as CS, AI, or Data Science, or in the humanities.
- English proficiency at B1 (intermediate) level or higher to draft project rubrics.
- Native-level fluency in at least one language.
- Strong writing skills and logic-based decision-making.
- Successful completion of a subject-specific qualification exam.
- ID verification as part of the hiring process.
Compensation
Expert pay rates range from $20 to $50 USD per hour, with payment terms defined for each project.
Community & Work Style
- Remote, flexible engagement.
- Part of a 1M+ global AI community of contributors and experts.
- Collaborate with specialists across 20+ domains and 500+ languages and dialects.
- Opportunity to contribute to AI research, evaluation, and fine-tuning workflows.
