Senior AI Engineer (Remote) — RLHF & Model Alignment

prolificacademicltd; United States; Remote; USD 83,000 - 110,000; Full time

Benefits offered by this job

Remote engagement

Flexible hours

Short skills assessment

Job summary

prolificacademic ltd seeks senior AI/ML engineers for remote, task-based research contributions. You will review model explanations, audit training loops, and provide structured feedback to advance RLHF and alignment. The role emphasizes deep technical evaluation and careful attention to bias and correctness.

Remote, flexible hours, candidates near San Francisco or willing to work US-time zones. Short entry assessment required; tasks typically fit within one hour sessions.

Qualifications

• Degree in a quantitative ML field with emphasis on ML research.
• Proven experience building, deploying, or fine-tuning ML models.
• Strong grasp of Transformers, CNNs, RNNs and optimization techniques.
• Experience with prompt engineering, RLHF or RAG workflows.
• Attention to detail for detecting hallucinations or data issues.

Responsibilities

• Evaluate explanations of model architectures, loss functions, and backpropagation.
• Audit ML code and notebooks, including training loops and evaluations.
• Provide structured feedback to support RLHF and model alignment.
• Assess how models reason on chain-of-thought prompts and identify failures.
• Run benchmarks comparing model outputs using technical taxonomies.

Skills

ML model evaluation

RLHF

Prompt engineering

Model alignment

Theory of ML proofs

Education

Bachelor's degree in CS/AI/Robotics

Master's or PhD in ML-focused field

Tools

PyTorch

TensorFlow/Keras

Weights & Biases

LangChain

Pinecone

Job description

prolificacademic ltd seeks senior AI/ML engineers for remote, task-based research contributions. You will review model explanations, audit training loops, and provide structured feedback to advance RLHF and alignment. The role emphasizes deep technical evaluation and careful attention to bias and correctness.

Remote, flexible hours, candidates near San Francisco or willing to work US-time zones. Short entry assessment required; tasks typically fit within one hour sessions.