[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"job-ai-annotation-work-from-home-germany":3},{"Ques":4,"Slug":22,"Header":23,"job_category":45},{"title":5,"content":6},"Frequently Asked Questions",[7,10,13,16,19],{"A":8,"Q":9},"Yes. This is a remote, full-time role, and you must be based in Germany.","Is this role remote?",{"A":11,"Q":12},"You will evaluate and rank model outputs, perform QA evaluation, write rationales and reasoning, validate against annotation guidelines compliance, and complete content safety labeling to support training data quality and model performance improvement.","What tasks will I do?",{"A":14,"Q":15},"AI or annotation experience is preferred but not required. You must be able to follow guidelines precisely and produce consistent large language model evaluation judgments.","Do I need AI experience?",{"A":17,"Q":18},"Fluency in English and German is required, as you will review and evaluate content in both languages.","What languages are required?",{"A":20,"Q":21},"Domains vary and can include general knowledge, customer-support style conversations, summarization, rewriting, safety-sensitive content, and other prompt evaluation scenarios used in RLHF and LLM evaluation workflows.","What domains are covered?","ai-annotation-work-from-home-germany",{"desc":24,"title":25,"content":26},"Rexzone is hiring Germany-based, bilingual English\u002FGerman AI Generalist Trainers to support RLHF and large language model evaluation by assessing, ranking, and QA-reviewing model outputs to improve training data quality and drive model performance improvement.","Germany-Based English & German AI Generalist Trainer 2026 May",[27,30,33,36,39,42],{"h2":28,"desc":29},"About the Role","As a Germany-based English & German AI Generalist Trainer at Rexzone, you will contribute to AI\u002FLLM workflows by performing RLHF-style evaluations, ranking responses, and writing clear rationales that support training data quality and model performance improvement. You will apply annotation guidelines compliance to validate model-generated outputs in English and German, flag content safety issues, and ensure consistent QA evaluation across tasks. This is a remote, full-time role focused on large language model evaluation and high-quality feedback loops that improve production-grade AI systems.",{"h2":31,"desc":32},"Key Responsibilities","Evaluate and rank model-generated outputs using RLHF and large language model evaluation criteria; Perform QA evaluation to verify accuracy, completeness, tone, and safety across English and German content; Write concise, evidence-based rationales and reasoning notes that explain rankings and drive model performance improvement; Validate task outputs against annotation guidelines compliance, including edge cases and ambiguity resolution; Label and review training data with strong attention to training data quality, consistency, and traceability; Perform prompt evaluation and response comparison, identifying failure modes and recommending improvements; Execute content safety labeling and escalation for policy-violating or sensitive material; Track errors, run validation checks, and propose updates to guidelines to reduce disagreement and improve throughput.",{"h2":34,"desc":35},"Basic Qualifications","Must be based in Germany and authorized to work as a remote contractor\u002Femployee as applicable; Fluency in English and German (written and reading comprehension required); Strong analytical skills with the ability to compare alternatives, identify subtle quality differences, and justify rankings; High attention to detail and consistency when following annotation guidelines and QA checklists; Ability to write clear rationales and structured reasoning under time constraints; Comfort working with web-based labeling tools and handling repetitive evaluation workflows.",{"h2":37,"desc":38},"Preferred Qualifications","Prior experience in data labeling, LLM evaluation, prompt evaluation, or QA evaluation; Familiarity with RLHF concepts, preference\u002Franking tasks, and model behavior analysis; Experience with content safety labeling or policy-based review; Self-driven, reliable, and able to work independently in a remote environment while meeting quality targets; Interest in improving training data quality and contributing to continuous model performance improvement.",{"h2":40,"desc":41},"Compensation and Schedule","Compensation is $35–$40 USD per hour, depending on assessment performance and ongoing quality metrics. Full-time availability is expected, with tasks delivered remotely and measured through throughput and annotation guidelines compliance.",{"h2":43,"desc":44},"How to Apply","Apply through Rexzone with an updated CV and a short note confirming Germany-based location and English\u002FGerman fluency. Selected candidates will complete a qualification assessment covering ranking, QA review, reasoning quality, and large language model evaluation criteria.","AI Data Operations"]