Research Engineer / Scientist

What you'll do

• Set strategy for training data and evaluation, including data acquisition, curation, synthetic generation, benchmarks, and release standards.
• Lead evaluation and data studies, including contamination controls, model-based judging, and product-aligned measures of user experience.
• Direct large-scale data-system architecture and lead technical standards, cross-team adoption, mentoring, and, where applicable, team management.

What they ask for

• PhD and 5+ years, or MS and 8+ years, in machine learning or data infrastructure, including leading technical work at scale.
• Deep expertise in LLMs or multimodal foundation models, with depth in evaluation methodology or large-scale training data systems and working fluency in the other.
• Track record of training-data or evaluation work on large models that measurably improved quality or changed how models are built.
• Hands-on proficiency in Python and an ML framework such as PyTorch or JAX.

Nice to have

• Experience with training data across model-training stages, multimodal data, source code, benchmarks, judging, contamination controls, or human evaluation.
• Experience with distributed data pipelines, data governance, or managing engineers or researchers.

Pay

$216,200 and $324,800 USD base salary

Tools

Python, PyTorch, JAX, Apache Beam, Spark

AI focus

Develops foundation models by leading training-data strategy and curation, evaluation, synthetic data generation, and systems that support model training and product-quality decisions.