Overview Remote, full-time engineering roles for mid-senior STEM graduates in the United States working on real AI/ML systems. You will help build and improve LLM training pipelines using RLHF, data labeling programs, and model evaluation/QA evaluation to drive measurable model performance improvement. What You Will Do
Design and ship ML components for LLM training pipelines
Partner with data operations to define annotation guidelines and labeling instructions
Build and iterate RLHF workflows (ranking, preference data, critique signals)
Run prompt evaluation and model evaluation to diagnose failure modes
Implement QA evaluation checks for annotation guidelines compliance
Coordinate NLP tasks (e.g., named entity recognition, classification) and CV annotation (bounding boxes, segmentation)
Contribute to content safety labeling policies and sampling strategies
Track improvements via offline metrics, error analysis, and dataset/versioning practices
Required Qualifications
STEM degree (or equivalent experience)
Strong Python and software engineering fundamentals
Hands-on ML experience in NLP and/or Computer Vision
Familiarity with RLHF, LLM evaluation, and/or prompt evaluation concepts
Ability to write clear specs for labeling and QA evaluation processes
Work Setup Remote (US). Collaborate asynchronously with distributed teams and may interface with AI labs, startups, and annotation vendors. Compensation Base pay range: $30–$50 per hour.