Principal Applied Scientist, AI Data Platform (CoreAI): Microsoft

This role is within Microsoft’s CoreAI team, which is building a central AI Data Platform. The team's mission is to break down Microsoft’s data silos and manage the full lifecycle of various data types (first-party, third-party, synthetic, and human-labeled) to accelerate AI model development with secure, reusable, and compliant datasets. As a Principal Applied Scientist, you will drive scientific innovation in data generation, validation, evaluation, and automation, setting the vision for intelligent, ML-driven services that manage the end-to-end data lifecycle.

Important Notes
This position requires passing the Microsoft Cloud background check upon hire and every two years thereafter.

Responsibilities
Define the scientific vision and roadmap for ML- and agent-driven automation of the dataset lifecycle (ingestion, validation, governance, etc.).

Lead the design and deployment of advanced ML pipelines for synthetic data generation, augmentation, and human-in-the-loop workflows.

Establish evaluation methodologies to measure dataset quality, coverage, and impact on large-scale model training.

Advance state-of-the-art methods for data-centric AI, including LLM-based evaluation, gap mining, and bias/fairness detection.

Mentor and grow a team of applied scientists, providing technical leadership.

Collaborate with engineering leaders to integrate research into scalable, production-ready platform services.

Influence Microsoft’s AI strategy by shaping best practices for data-driven model development.

Requirements
Required Qualifications:

A Bachelor's degree in a relevant field (Statistics, CS, Engineering) with 6+ years of experience, OR a Master's degree with 4+ years of experience, OR a Doctorate with 3+ years of experience.

Experience applying ML/AI to real-world problems and leading applied science projects from concept to production.

Programming experience in Python and ML frameworks.

Demonstrated expertise in data quality, dataset evaluation, or synthetic data generation.

Preferred Qualifications:

A PhD in a relevant field.

Experience with LLMs, data-centric AI approaches, or intelligent agent-based systems.

Knowledge of data privacy, compliance, and governance in large-scale AI systems.

Familiarity with distributed data systems (e.g., Spark, Databricks, Azure Data Lake).

A strong publication record in top ML/AI conferences (e.g., NeurIPS, ICML, ICLR).