About The Role

Rex.zone is hiring Germany-based bilingual (English/German) AI Generalist Trainers to support RLHF and large language model evaluation. You will evaluate, rank, and QA model-generated responses, write clear rationales, and ensure annotation guidelines compliance to improve training data quality and model performance.

Key Responsibilities

Evaluate and rank model outputs against rubrics
Perform QA evaluation for consistency, safety, and instruction-following
Write concise rationales explaining reasoning and tradeoffs
Validate edge cases, escalate guideline gaps, and enforce annotation guidelines compliance
Execute data labeling and prompt evaluation with high precision
Run quality checks to improve training data quality
Support content safety labeling and policy-based decisions
Track errors and propose fixes that improve model performance
Maintain clear bilingual (English/German) documentation

Basic Qualifications

Based in Germany (remote, full-time)
Fluent in English and German (C1+ or equivalent)
Strong analytical skills and structured reasoning
Excellent attention to detail and consistency on repeated evaluations
Comfortable following rubrics and annotation guidelines
Ability to work independently with reliable turnaround

Preferred Qualifications

Experience in AI data labeling, RLHF, LLM evaluation, or QA evaluation
Familiarity with helpfulness, harmlessness, factuality, and instruction-following
Experience writing high-quality rationales and error analyses

Compensation

USD $30–$50 per hour (full-time, remote, Germany-based).

How To Apply

Apply with your resume/CV and a short note describing your experience with evaluation, QA, bilingual writing, and any exposure to RLHF or training data quality work.

Selected candidates will complete a skills assessment focused on large language model evaluation and annotation guidelines compliance.

Similar jobs