Member of Technical Staff - Post-Training | Microsoft Careers

What you'll do

• Design high-quality datasets and benchmarks for training AI models.
• Apply expertise in pre-training, post-training, and reinforcement learning for models.
• Develop scalable data pipelines for large, complex multimodal datasets.
• Assess real-world multimodal datasets for quality, diversity, and relevance.
• Build tools for dataset auditing, visualization, and versioning to streamline experiments.

What we're looking for

• Bachelor's Degree or equivalent experience with 3+ months of research internship.
• Master's Degree and 1+ year of related research experience preferred.
• Proficiency in Python and modern data libraries (Pandas, NumPy).
• Experience with distributed data frameworks like Spark and cloud ecosystems.
• Hands-on work with large-scale unstructured datasets including images, video, audio, and code.
• Proven ability to train AI models at significant scale.