Senior Data Labeling Jobs in Barcelona

Senior data labeling jobs in Barcelona at Rex.zone focus on training data quality for AI systems across NLP, computer vision, and LLM training pipelines. You will lead annotation workflows, refine annotation guidelines compliance, and run QA evaluation to drive model performance improvement. This remote, full-time role supports RLHF, prompt evaluation, named entity recognition, content safety labeling, and multimodal dataset curation used by AI labs, tech startups, and annotation vendors. If you build reliable ground truth, manage edge cases, and improve reviewer consistency, explore and apply through Rex.zone to help ship safer, higher-accuracy models.

Job Overview

Title: Senior Data Labeling Specialist Barcelona Date: 25-02-2026 Company: Rex.zone Country: US Remote Type: Remote Employment Type: FULL_TIME Experience Level: Mid-Senior Industry: Technology Job Function: Engineering Skills: data labeling, training data quality, annotation guidelines compliance, QA evaluation, RLHF, prompt evaluation, named entity recognition, computer vision annotation, content safety labeling, LLM training pipelines Salary Currency: USD Salary Min: 63360 Salary Max: 126720 Pay Period: YEAR

About the Role

You will lead senior-level data labeling and evaluation workflows for ML datasets used in large language model evaluation and multimodal model training. This includes building and enforcing annotation guidelines, resolving ambiguous edge cases, calibrating annotators, and improving inter-annotator agreement. You will partner with engineering and research stakeholders to translate model goals into labeling taxonomies, run sampling-based QA evaluation, and deliver training data quality improvements that measurably impact offline metrics and online model behavior.

Key Responsibilities

Own annotation project setup: taxonomy design, label definitions, and gold-standard examples for consistent ground truth creation Lead RLHF support tasks, including preference ranking and prompt evaluation aligned to safety, helpfulness, and factuality rubrics Drive QA evaluation programs: audits, double-pass review, error categorization, and root-cause analysis for systematic issues Support NLP labeling tasks such as named entity recognition, entity linking, classification, and retrieval relevance judgments Support computer vision annotation workflows such as bounding boxes, polygons, segmentation masks, and attribute tagging when needed Operationalize content safety labeling guidelines for policy, toxicity, self-harm, and sensitive content categories Report training data quality KPIs, reviewer consistency metrics, and model performance improvement insights to stakeholders Coordinate with annotation vendors or BPO teams when applicable; ensure throughput and accuracy targets are met Continuously iterate guidelines and tooling feedback to reduce ambiguity and improve labeling efficiency

Required Qualifications

Mid-Senior experience in data labeling, data annotation, or ML data operations supporting production AI/ML systems Strong understanding of annotation guidelines compliance, rubric design, and quality assurance methods Hands-on experience with QA evaluation workflows such as audits, adjudication, and calibration sessions Familiarity with LLM training pipelines, including RLHF concepts, preference data, and prompt evaluation best practices Ability to reason about ambiguity, write clear labeling instructions, and handle long-tail edge cases consistently Comfort collaborating cross-functionally with engineering, applied ML, product, and operations teams

Preferred Qualifications

Experience with named entity recognition and other structured NLP labeling tasks Experience with computer vision annotation, including segmentation and attribute tagging Experience labeling content safety datasets and interpreting policy-driven rubrics Experience with evaluation design for large language model evaluation and reviewer agreement measurement Exposure to annotation tooling, workflow automation, or dataset versioning practices

What Success Looks Like

Training data quality improves through fewer guideline violations, lower rework rates, and higher reviewer agreement Annotation guidelines are clear, testable, and resilient to edge cases across domains like NLP and computer vision QA evaluation identifies systematic failure modes and closes the loop with actionable guideline updates RLHF and prompt evaluation outputs are consistent, well-calibrated, and useful for model performance improvement Stakeholders can trace dataset changes to measurable improvements in model evaluation and production behavior

Work Modality and Role Fit

Remote: Yes, this role remains Remote Full-time: Yes, FULL_TIME employment type Also relevant to searches for: remote, full-time, contract, freelance, entry-level, and senior data labeling roles across AI labs, tech startups, BPOs, and annotation vendors Domains supported: NLP, computer vision, content safety, and LLM training pipelines

How to Apply on Rex.zone

Prepare a short summary of your data labeling and QA evaluation experience, including guideline design and audit methods List domains you have supported: RLHF, prompt evaluation, named entity recognition, computer vision annotation, or content safety labeling Apply through Rex.zone and be ready to discuss training data quality, edge-case handling, and reviewer calibration approaches

Frequently Asked Questions

Q: Is this a remote role even though the keyword includes Barcelona?
Yes. The role is explicitly marked Remote. The Barcelona keyword aligns with search intent for senior data labeling jobs tied to Barcelona, while the work arrangement remains Remote.
Q: What types of data labeling tasks will I lead?
You will lead data labeling and evaluation across RLHF preference data, prompt evaluation, named entity recognition, computer vision annotation, and content safety labeling, with an emphasis on training data quality and annotation guidelines compliance.
Q: How is quality measured in this role?
Quality is measured through QA evaluation methods such as sampling audits, adjudication, inter-annotator agreement, guideline violation rates, and error taxonomy tracking tied to model performance improvement.
Q: What is the difference between RLHF tasks and standard labeling?
Standard labeling often focuses on ground truth classification or structured tags, while RLHF work frequently involves preference ranking, rubric-based judgments, and prompt evaluation designed to shape model behavior in LLM training pipelines.
Q: Which employer types can this work support?
The workflows are commonly used by AI labs, tech startups, BPOs, and annotation vendors building datasets for NLP, computer vision, content safety, and large language model evaluation.
Q: What skills should I highlight for a senior data labeling role?
Highlight training data quality ownership, annotation guidelines compliance, QA evaluation, adjudication and calibration, edge-case reasoning, RLHF and prompt evaluation familiarity, named entity recognition knowledge, and experience coordinating multi-reviewer workflows.

230+Domains Covered

120K+PhD, Specialist, Experts Onboarded

50+Countries Represented

Industry-Leading Compensation

We believe exceptional intelligence deserves exceptional pay. Our platform consistently offers rates above the industry average, rewarding experts for their true value and real impact on frontier AI. Here, your expertise isn't just appreciated - it's properly compensated.

Work Remotely, Work Freely

No office. No commute. No constraints. Our fully remote workflow gives experts complete flexibility to work at their own pace, from any country, any time zone. You focus on meaningful tasks - we handle the rest.

Respect at the Core of Everything

AI trainers are the heart of our company. We treat every expert with trust, humanity, and genuine appreciation. From personalized support to transparent communication, we build long-term relationships rooted in respect and care.

Ready to Shape the Future of AI Data Operations?

Apply Now.