Pretraining Data
Core
Design, curate, and evaluate large-scale datasets to train frontier AI models at Microsoft AI.
Role type
Senior Research Engineer (Pre-training Data)
Builds
High-quality datasets and scalable data pipelines for foundation models
Domain
Artificial Intelligence / Deep Learning / Data Engineering
Deliverable
production ML models
Required skills
Python, Pandas, NumPy, Statistics, Exploratory Data Analysis, Data Processing Frameworks (Spark, Ray, Apache Beam), Large-scale Data Engineering
Preferred skills
Advanced Statistics, Data Analysis, Data Curation Strategies, Ethical AI Alignment
Technologies
Python, Pandas, NumPy, Spark, Ray, Apache Beam
Responsibilities
Design and curate high-quality datasets for frontier AI models; Develop and maintain scalable data pipelines for ingestion, processing, filtering, and annotation; Analyze real-world datasets to assess quality, diversity, and relevance; Build lightweight tools for dataset auditing, visualization, and versioning; Collaborate with Safety, Ethics, and Governance teams to ensure responsible AI practices.