Alpheratz Project - Catalan (Spain) Translation Quality Rater

Welo is an AI services company that partners with leading organizations to create high-quality training data and evaluation frameworks for machine learning systems. The company works across multiple languages and specialized domains to ensure AI models perform accurately in real-world applications.

Welo's Data team specializes in linguistic quality assurance and model improvement through human expertise. They collaborate with native speakers across the globe to refine machine translation systems and enhance customer-facing content quality. This role sits at the intersection of language mastery and AI development, where your judgment directly shapes how models learn to produce better outputs.

We're seeking a native Catalan speaker to evaluate and refine machine-translated customer service materials for the Alpheratz project. This is remote contract work based in Spain, with flexible scheduling around your availability. Tasks arrive on an as-needed basis throughout the year, typically requiring one to two hours per week, though volume fluctuates based on project demands.

What you'll do

• Review and edit machine-translated customer service content to ensure accuracy, consistency, and alignment with brand voice and terminology standards.

• Assess the quality of machine translation output using established severity rating systems and provide detailed feedback on errors and limitations.

• Create high-quality human translations that serve as reference data for model training and quality benchmarking.

• Apply provided style guides, tone guidelines, and terminology lists to maintain standardization across all reviewed materials.

• Document patterns in machine translation failures and communicate recurring quality issues to the project team.

What you'll bring

• Native or near-native fluency in Catalan (Spain) with professional proficiency in English.

• Demonstrated experience in translation, localization, post-editing of machine output, or similar language work.

• Knowledge of how machine translation systems work and ability to identify specific types of translation errors and quality gaps.

• Strong ability to catch subtle linguistic issues and maintain consistency across large batches of content.

• Comfort working independently in a remote setting and following detailed project specifications.

Nice to have

• Background in evaluating or rating machine-generated text.

• Experience with translation management tools or quality assurance platforms.

What they offer

• Hourly rate of $21.90.

• Fully remote contract position allowing flexible scheduling.

• Intermittent work model that fits around other commitments.