Senior Applied Scientist

Responsibilities

• Drive projects from initial design through implementation, experimentation, and final shipping to users.
• Perform deep dives into large-scale data to identify gaps, develop heuristics, and implement solutions.
• Utilize LLMs to create high-quality model and evaluation prompts.
• Design and implement engineering pipelines and infrastructure for model training and evaluation.
• Develop advanced evaluation techniques, datasets, criteria, and metrics for state-of-the-art models.
• Own the end-to-end fine-tuning process, including dataset creation, filtering, review, and continuous iteration.
• Collaborate within a diverse, geographically distributed team to impact the design and scalability of Microsoft products.

Requirements

• Bachelor's Degree in Statistics, Computer Science, Electrical/Computer Engineering, or a related field with 7+ years of experience (or equivalent Master's/Doctorate with relevant experience).
• Proven experience in large-scale data analysis to identify patterns or create evaluation datasets.
• Strong proficiency with common machine learning and deep learning frameworks.
• Hands-on experience with Large Language Models (LLMs) and advanced prompting techniques.
• Ability to communicate complex technical details clearly across organizational boundaries.

Preferred Qualifications

• Experience with PyTorch or TensorFlow.
• 3+ years of experience in research, including creating publications (patents, peer-reviewed papers, or libraries).
• Experience presenting research at industry conferences or academic events.
• 1+ years of experience developing and deploying live production systems within a product team.
• Prior experience applying deep learning to Search, Recommendation, NLP, or Document Understanding.
• Experience with Azure or other cloud-based execution graphs and pipelines.
• Solid track record of publications at top-tier conferences such as ACL, EMNLP, NIPS, or AAAI.

About the Company

Microsoft's Turing team is an innovative engineering and applied research group working on state-of-the-art deep learning models and pioneering conversational search experiences. The team spearheads innovation for core Copilot experiences across the Microsoft ecosystem, including Office and Windows.