Applied Scientist

@ Microsoft

Overview

The Applied Scientist position within Microsoft M365 Core is available at multiple Development Centers located in the UK, Estonia, Czech Republic, and Romania. As an Applied Scientist, you will contribute to the advancement of artificial intelligence, machine learning, and model optimization by developing and fine-tuning large language models (LLMs) and task-specific agents. This role focuses on delivering intelligent features across the Microsoft 365 ecosystem such as Copilot, Outlook, Teams, SharePoint, and OneDrive.

Responsibilities

• Conduct research and development to enhance AI model training and evaluation.
• Fine-tune LLMs on tenant data for task-specific agents.
• Collaborate with engineers and researchers on deep learning and NLP challenges.
• Measure, analyze, and report on model performance using statistical methodologies.
• Foster a culture of inclusion, innovation, and accountability.

Qualifications

Candidates should have a Bachelor’s, Master’s, or Doctorate degree in Statistics, Econometrics, Computer Science, Electrical/Computer Engineering, or a related field. Demonstrated experience in training and fine-tuning AI/ML models, particularly LLMs, and proficiency in at least one programming language like Python is required. Experience in productization or shipping ML/AI components at scale is essential.

Preferred Qualifications

Additional merits include creating publications (e.g., patents, peer-reviewed papers), managing end-to-end technical solutions, and building generative AI pipelines with approaches such as RAG (Retrieval Augmented Generation).

Team Culture

This role is embedded in a research-driven environment that values deep thinking, technological innovation, and rapid execution within an inclusive and growth-oriented culture at Microsoft.

Key skills/competency

• Artificial Intelligence
• Machine Learning
• Large Language Models
• Natural Language Processing
• Deep Learning
• Model Optimization
• Predictive Analytics
• Statistical Methodologies
• Research
• Python Programming