STEM Careers in the United States

STEM Careers in the United States (Remote, Full-Time)

Rex.zone is hiring a US‑based, STEM‑aligned engineering professional to help build and evaluate real AI/ML systems with a focus on large language model (LLM) workflows. You will translate product goals into measurable evaluation signals, training data quality controls, and reliable tooling that improves model performance across iterations.

What You Will Do

• Build and improve tooling for LLM evaluation, including prompt evaluation and QA evaluation loops
• Support RLHF evaluation feedback cycles and rubric‑based scoring
• Define and maintain annotation guidelines and automated compliance checks
• Design sampling strategies, acceptance criteria, and drift checks for training data quality
• Analyze error patterns and propose fixes that drive model performance improvement
• Partner with labeling teams/vendors on ambiguity resolution, rework reduction, and throughput/quality reporting
• Implement dashboards and metrics (e.g., inter‑annotator agreement, precision/recall proxies)
• Help enforce privacy, security, and policy constraints for sensitive datasets

Required Qualifications

• Mid‑senior experience in engineering, computer science, data science, or a related STEM field
• Strong programming fundamentals and ability to ship production‑quality tools
• Familiarity with ML concepts and LLM evaluation methods
• Experience with data labeling, QA evaluation, or dataset curation for NLP and/or computer vision
• Ability to write clear specs, rubrics, and annotation guidelines
• Strong analytical skills for debugging dataset and evaluation issues in a remote, cross‑functional environment

Preferred Qualifications

• Experience with RLHF or preference modeling workflows
• Hands‑on prompt evaluation and rubric‑based scoring
• Named entity recognition (NER), taxonomy/ontology design, or content safety labeling
• Vendor/BPO collaboration for large‑scale annotation operations

Compensation

Hourly base salary range: $30–$50 per hour (USD).

Remote, full‑time role based in the United States.