Das ist der Job
Review multi-step agent workflows, including screenshots and reasoning traces, to assess accuracy and completeness.
Darum lohnt es sich
Type: Hourly contract
Compensation: $20–$30/hour
Location: Remote
Commitment: 10–40 hours/week
Role Responsibilities
Evaluate outputs from large language models and autonomous agent systems using defined rubrics and quality standards. Apply benchmarking criteria consistently while identifying edge cases and recurring failure patterns.
Provide structured, actionable feedback to support model refinement and product improvements. Participate in calibration sessions to ensure consistent evaluation alignment across reviewers. Adapt to evolving guidelines and ambiguous scenarios with sound judgment. Document findings clearly and communicate insights to relevant stakeholders.
Requirements
Strong experience in LLM evaluation, AI output analysis, QA/testing, UX research, or similar analytical roles. Proficiency in rubric-based scoring, benchmarking frameworks, and AI quality assessment. Excellent attention to detail with strong decision-making skills in ambiguous cases.
Proficient English communication skills (written and verbal). Ability to work independently in a remote environment. Comfortable committing to structured eva
