Senior AI/ML Engineer - Remote, Task-Based Evaluations

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

Benefits offered by this job

Remote work

Flexible scheduling

Job summary

prolificacademicltd is seeking senior AI/ML engineers for remote, task-based technical work that trains and evaluates large language models. You will critique model behavior, code, and reasoning, focusing on deep technical feedback rather than ongoing product development.

Responsibilities include auditing training loops, evaluating explanations, and contributing to RLHF safety and alignment efforts. A strong foundation in neural networks and prompt engineering is essential.

Qualifications

• Advanced degree in a quantitative ML field or equivalent research experience.
• Production experience with ML models in training, deployment, or fine-tuning.
• Familiar with neural architectures (Transformers, CNNs, RNNs) and optimization methods.

Responsibilities

• Assess AI-generated explanations for architectural and mathematical accuracy.
• Audit ML code, notebooks, training loops, and preprocessing for correctness and efficiency.
• Provide expert feedback to strengthen RLHF processes for safety and helpfulness.
• Analyze chain-of-thought reasoning to identify failures or flawed logic.
• Benchmark model outputs against technical criteria.

Skills

Prompt engineering

RLHF

RAG workflows

Model evaluation

ML theory

Education

BS/MS/PhD in CS/AI/Robotics or related ML

Tools

PyTorch

TensorFlow/Keras

NumPy

Pandas

Scikit-learn

Hugging Face Transformers

Job description

prolificacademicltd is seeking senior AI/ML engineers for remote, task-based technical work that trains and evaluates large language models. You will critique model behavior, code, and reasoning, focusing on deep technical feedback rather than ongoing product development.

Responsibilities include auditing training loops, evaluating explanations, and contributing to RLHF safety and alignment efforts. A strong foundation in neural networks and prompt engineering is essential.