Key Responsibilities: 1. Translation Accuracy & Cultural Adaptation Identify semantic errors and idiomatic misuses (e.g., contextual suitability of Mexican Spanish phrases like "¡Qué chévere!") in EN-DE/EN-TR/EN-ES translations from a native-speaker perspective. Ensure localized accuracy for marketing emails, product descriptions, and UI copy. Resolve technical translation discrepancies (e.g., mechanical part terminology, e-commerce platform UI terms) and build domain-specific glossaries. 2. AI Model Evaluation & Fine-Tuning Systematically assess LLM outputs using metrics (BLEU scores, human readability) and tag error patterns (e.g., passive voice overuse, gender-agreement errors) to guide model iteration. Train models with high-quality bilingual pairs (e.g., comparing human vs. AI translations) to enhance semantic understanding of business contexts (inquiries, contracts, etc.). 3. Cross-Functional Collaboration Align AI standards with e-commerce user journeys (search → checkout → post-sale) to optimize real-time translation performance (e.g., cart/checkout page responsiveness). Coordinate with UX designers on post-translation constraints (e.g., line-break handling for German compound words) to ensure technical feasibility and user experience consistency. 4. Data-Driven Optimization Diagnose systematic translation errors (e.g., missing Turkish case suffixes) and develop scalable solutions to accelerate model training. Quantify model improvement (e.g., error rate reduction) through version-by-version quality benchmarking to support decision-making.

Minimum 2 years of professional experience in AI model training, linguistic evaluation, or multilingual content strategy Native or near-native proficiency in English plus German/Turkish/Mexican Spanish Demonstrated expertise in localization quality assessment and cultural adaptation Familiarity with NLP evaluation metrics (BLEU, TER, etc.) and common translation tools Strong analytical skills to identify linguistic patterns and optimize model performance