Germany-Based English & German AI Generalist Trainer 2026 May

Rex.zone; United States; On-site; USD 73,000 - 101,000; Full time

Job summary

Rex.zone is hiring Germany-based AI Generalist Trainers to support large language model evaluation through RLHF, LLM evaluation, and training data quality workflows. You will assess, rank, and QA model outputs in English and German, write clear rationales, and follow annotation guidelines to drive model performance.

As a Germany-based remote role, you must be fluent in German and English, with strong writing, analytical skills, and meticulous attention to guidelines.

Qualifications

• Fluent in German and English with strong writing skills.
• Analytical mindset with attention to detail.
• Ability to follow complex annotation guidelines.
• Comfortable with web tools and structured review workflows.

Responsibilities

• Evaluate model responses for correctness and usefulness in English and German.
• Rank outputs using RLHF-style criteria.
• Perform QA checks to ensure data quality and guideline compliance.
• Explain reasoning with concise rationales for rankings and edits.
• Identify ambiguity and propose clarifications to guidelines.
• Review prompts and responses for policy alignment and content safety labeling when required.
• Track errors and feedback to support model performance improvements.

Skills

German language

English language

Analytical thinking

Attention to detail

Job description

About Rex.zone

Rex.zone is hiring Germany-based AI Generalist Trainers to support large language model evaluation through RLHF, LLM evaluation, and training data quality workflows. You will assess, rank, and QA model outputs in English and German, write clear rationales, and follow annotation guidelines compliance to drive model performance improvement.

About The Role

As a Germany-based English & German AI Generalist Trainer at Rex.zone, you will improve AI systems by performing large language model evaluation across diverse domains including general knowledge, customer-support style tasks, summarization, rewriting, reasoning, and content safety labeling.

Responsibilities

• Evaluate model-generated responses for correctness, completeness, and helpfulness in English and German
• Rank multiple outputs using RLHF-style preference criteria and rubric-driven scoring
• Perform QA evaluation and validation checks to ensure training data quality and annotation guidelines compliance
• Write concise, well-structured rationales explaining reasoning behind rankings and edits
• Identify ambiguity/edge cases, elevate guideline gaps, and propose clarifications
• Review prompts and responses for policy alignment, including content safety labeling when required
• Track errors, patterns, and feedback to support continuous model performance improvement

Basic Qualifications

• Based in Germany (remote role) and eligible to work
• Fluent in German and English with strong writing skills in both languages
• Strong analytical skills, attention to detail, and consistency following guidelines
• Comfortable with web tools, spreadsheets, and structured review workflows

Preferred Qualifications

• Experience with data labeling, prompt evaluation, QA evaluation, or LLM evaluation
• Familiarity with RLHF concepts, rubric-based scoring, and training data quality best practices
• Experience with editorial review or policy-aligned content review