R

Entry Level STEM Jobs United States

Accepting applications

Rex.zone · United States

Full-Time Mid_senior AIPythonaiarmasic
Posted
3 May
Category
Test
Experience
Mid_senior
Country
United States
Entry Level STEM Jobs United States (Remote, US)

Support AI/ML engineering-adjacent workflows that improve LLM training pipelines through high-quality data labeling, RLHF-style evaluation, QA evaluation, and prompt evaluation. You will apply annotation guidelines compliance and structured documentation to strengthen training data quality and model performance.

What You Will Do

Execute data labeling across text, image, and multimodal datasets
Perform QA evaluation using sampling plans, disagreement analysis, and error taxonomies
Complete RLHF preference ranking and prompt evaluation aligned to rubrics
Annotate NLP tasks (e.g., named entity recognition, intent classification) with consistent schemas
Support computer vision annotation (bounding boxes, segmentation) as needed
Apply content safety labeling policies for harmful or sensitive content
Document edge cases and provide rationales to improve guidelines and training data quality
Collaborate asynchronously with distributed engineering and ops partners

Required Qualifications

US-based and available for remote, full-time work
STEM background (degree, bootcamp, or equivalent experience) and quantitative reasoning
Ability to learn and follow detailed annotation guidelines
Strong written communication and attention to detail
Familiarity with spreadsheets and basic Python or SQL for lightweight data checks

Preferred Qualifications

Experience with LLM evaluation, RLHF, or prompt evaluation
Exposure to NLP and/or computer vision annotation tools
Understanding of content safety labeling and policy interpretation
Experience with training data quality metrics or inter-annotator agreement

Compensation

Hourly base pay range: $30–$50/hr.

How To Apply

Apply via Rex.zone and highlight STEM projects, Python/SQL, and any experience with data labeling, QA evaluation, RLHF, prompt evaluation, or annotation guidelines compliance.
Show more Show less