Principal Applied Scientist

Do you aspire to be part of a dynamic team that develops cutting-edge products and machine learning solutions at Microsoft, reaching millions of users every month The Microsoft Turing team is an innovative group specializing in engineering and applied research, dedicated to advancing deep learning models, large language models, and groundbreaking conversational search experiences. At the forefront of conversational search platform and innovation, the team drives core copilot experiences within Microsoft's ecosystem, spanning BizChat, Office, and Windows. As a Principal Applied Scientist within the Turing team, you will lead and execute various data science tasks within tight timelines. Your responsibilities will include hands-on activities such as model training, evaluation set creation, infrastructure development for training and evaluation processes, and more. Collaboration with internal and external data science, product, and engineering teams across different time zones will be essential for successful project delivery. Microsoft's mission revolves around empowering individuals and organizations worldwide to accomplish more. As team members, we embrace a growth mindset, strive for innovation to empower others, and foster collaboration to achieve our collective objectives. Upholding values of respect, integrity, and accountability, we cultivate an inclusive culture where everyone can excel professionally and personally. In your role as a Principal Applied Scientist, you will: - Drive projects from inception to implementation, engaging in data analysis, heuristic formulation, model creation using Large Language Models (LLMs), and establishing engineering pipelines for model execution. - Provide documentation, guidance to junior team members, and collaborate with stakeholders across different time zones to ensure project alignment and timely progress. - Develop evaluation techniques, datasets, criteria, and metrics for model assessments, often involving State-of-the-Art (SOTA) models and metrics/datasets. - Engage in hands-on tasks such as pre-training, fine-tuning, and utilization of language models, encompassing dataset preparation, review, and continual refinement. Proficiency in training frameworks, formats, and stacks like megatron is also required. This role demands active participation in a diverse, globally dispersed team environment that values collaboration and innovation. You will play a pivotal role in shaping the design, functionality, security, performance, scalability, manageability, and supportability of Microsoft products leveraging our deep learning technology. Qualifications: Required Qualifications: - Bachelor's, Master's, or Doctorate degree in Statistics, Econometrics, Computer Science, Electrical/Computer Engineering, or related field with 8+ to 12+ years of relevant experience. - At least 3 years of experience in delivering team-level outcomes. - 2+ years of industrial coding experience in languages like C++, C#, C, Java, or Python. - Previous exposure to data analysis in large-scale systems, pattern identification, or evaluation dataset creation. - Familiarity with machine learning, deep learning frameworks, Large Language Models (LLMs), and prompting techniques. - Strong communication skills to convey technical details effectively across organizational boundaries. Preferred Qualifications: - 5+ years of experience in creating publications such as patents, libraries, or peer-reviewed academic papers. - 2+ years of experience in presenting at conferences or industry events as an invited speaker. #MSAI #Turing #LLMs #Modeltraining #M365CORE #MSAI #Turing,