Remote Data Annotation Jobs in Paris
Rex.zone
LinkedInGeneralGeneralbothRemote
LocationParis, Île-de-France, France
CountryFrance
Work modeRemote
LevelGeneral
FunctionGeneral
Fieldboth
Employment typeGeneral
Posted20/08/2026
Job description
Remote Data Annotation Jobs in Paris (Full-Time, Remote) Rex.zone supports real AI/ML training workflows by turning raw text, images, audio, and video into high-quality labeled datasets for large language models and computer vision systems. You will follow annotation guidelines, perform QA evaluation, and contribute to RLHF and prompt evaluation tasks that improve model performance, training data quality, and content safety outcomes. What You Will Do Deliver high-precision annotations for NLP tasks (e.g., named entity recognition, text classification, relevance judgments) Perform RLHF-style preference comparisons, rubric-based LLM evaluation, and prompt evaluation Annotate computer vision data (bounding boxes, polygons, segmentation masks) when required Apply content safety labeling using consistent taxonomy and policy categories Run QA evaluation routines: spot checks, disagreement analysis, error categorization, and corrective actions Document edge cases, label definitions, and decision rationales for auditability and consistency at scale Requirements Mid-Senior experience in data annotation/data labeling or QA evaluation for AI/ML datasets Strong ability to interpret guidelines and resolve ambiguity consistently Familiarity with RLHF concepts, preference ranking, and LLM evaluation rubrics Understanding of NLP and/or computer vision annotation workflows and common error modes High attention to detail and strong written communication for decision logs How Success Is Measured Annotation accuracy, guideline compliance, and inter-annotator agreement signals Reduced error rates through clear edge-case handling and actionable QA analysis Reliable throughput while maintaining quality across labeling queues Apply Submit your profile on Rex.zone and complete any required calibration tasks mirroring production annotation and LLM evaluation workflows.