Remote AI Engineer for Model Alignment & Benchmarking

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

Benefits offered by this job

Up to $80 per hour

Fully remote work

Flexible scheduling

Job summary

Prolific Academic Ltd. seeks senior AI/ML engineers to train and evaluate cutting-edge language models through paid research tasks. Work remotely with flexible scheduling, auditing model logic and data quality, and providing detailed human feedback for RLHF pipelines.

The role focuses on evaluating prompts, improving alignment, and benchmarking model outputs with expert judgment from experienced Seattle-area engineers. Remote setup with a short onboarding assessment.

Qualifications

• Degree in CS/AI/robotics or related quantitative field with ML concentration.
• Production ML experience: building, deploying or fine-tuning models.
• Knowledge of transformers, CNNs, RNNs and optimization techniques.
• Familiarity with RLHF, prompt engineering and RAG workflows.
• Detail-oriented in auditing model logic and data quality.

Responsibilities

• Review explanations of model architectures, loss functions, and backpropagation.
• Validate ML code and notebooks, including training loops and data prep.
• Provide high-quality human feedback for RLHF pipelines.
• Evaluate model responses and identify reasoning breakdowns in prompts.
• Benchmark outputs across defined metrics and taxonomies.

Skills

Attention to detail

Prompt engineering

Model evaluation

Analytical thinking

Communication

Education

BS/MS/PhD in Computer Science or related field

Tools

PyTorch

TensorFlow/Keras

NumPy

Pandas

Scikit-learn

Hugging Face Transformers

AWS SageMaker

LangChain

Pinecone/Milvus/Weaviate

Job description

Prolific Academic Ltd. seeks senior AI/ML engineers to train and evaluate cutting-edge language models through paid research tasks. Work remotely with flexible scheduling, auditing model logic and data quality, and providing detailed human feedback for RLHF pipelines.

The role focuses on evaluating prompts, improving alignment, and benchmarking model outputs with expert judgment from experienced Seattle-area engineers. Remote setup with a short onboarding assessment.