German AI Data & RLHF Linguist

As agentic AI platforms expand into dozens of new language markets, native German-language data is a defined gap for global foundation-model teams. In this fully remote role, you'll split your time between text/multimodal annotation and RLHF preference labeling for German-language AI programs, working to versioned taxonomies and rubrics in Jwuma.

Key responsibilities

• Annotate German-language text, image, or multimodal data to project-specific taxonomy
• Compare and rank German-language model outputs against a rubric to produce RLHF preference data
• Write concise rationale notes for RLHF judgments and flag edge cases for QA leads
• Peer-review teammates' annotation batches ahead of QA handoff
• Maintain 97%+ accuracy against gold-standard benchmark sets

What we look for

• Native or near-native German with strong professional English
• Strong, consistent judgment for both structured annotation and RLHF preference comparisons
• Reliable internet and a self-managed remote work setup
• Eligible to work remotely in Germany or another EU/EEA country
• Comfortable learning new tools and taxonomies (full paid training provided)

Nice to have

• Background in linguistics, translation, journalism, or research
• Prior annotation, content-moderation, or QA experience
• Familiarity with generative AI/chatbot tools as a user

Compensation and benefits

• Competitive hourly contract pay, paid via Jwuma timesheets
• Fully remote across Germany/EU with flexible scheduling
• Paid onboarding on taxonomy, rubric, and platform tools
• GDPR-aligned data handling on every project

Common questions