Senior AI/ML Research Contributor - Remote, Flexible Tasks

prolificacademicltd; United States; Remote; USD 91,000 - 129,000; Part time

Job summary

prolificacademicltd seeks seasoned AI/ML engineers to contribute to training and evaluation of large language models. This is a remote, flexible, task-based engagement rather than a traditional role.

Senior engineers with deep ML expertise can dedicate focused blocks of time for technical review work, ideally based in or near Sacramento. Responsibilities include auditing model explanations, ML code, RLHF feedback, and benchmarking model outputs.

Qualifications

• Bachelors/Masters/PhD in CS/AI/Robotics with ML focus.
• Hands-on experience deploying or fine-tuning ML models in production.
• Experience with neural nets: Transformers, CNNs, RNNs.
• Experience with prompt engineering and RLHF workflows.
• Attention to detail for hallucinations, biases, or logical failures.
• Ability to review ML proofs and assess data contamination.

Responsibilities

• Assess explanations of model architectures and loss functions for soundness.
• Audit ML code, notebooks, training loops, and evaluation pipelines.
• Provide structured human feedback for RLHF pipelines.
• Evaluate reasoning on chain-of-thought prompts and breakdowns.
• Run benchmarks across model outputs using defined metrics.

Skills

Machine learning

Transformers

RNNs

CNNs

Prompt engineering

RLHF workflows

Proofs review

Data contamination

Production ML

Education

Bachelor's degree

Master's degree

PhD

Tools

PyTorch

TensorFlow

NumPy

Pandas

Scikit-learn

Hugging Face

Weights & Biases

LangChain

SageMaker

Vertex AI

Pinecone

Milvus

Weaviate

Job description

prolificacademicltd seeks seasoned AI/ML engineers to contribute to training and evaluation of large language models. This is a remote, flexible, task-based engagement rather than a traditional role.

Senior engineers with deep ML expertise can dedicate focused blocks of time for technical review work, ideally based in or near Sacramento. Responsibilities include auditing model explanations, ML code, RLHF feedback, and benchmarking model outputs.