Microsoft Applied Scientist Microsoft AI – PhD Redmond 

Taiwan, Taoyuan City 

536228283

Please note this application is only for roles based in our Redmond, Washington office. For roles in other offices in the United States, please see our .

Required Qualifications:

• Doctorate (or currently pursuing a doctorate) in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field.
• 3+ years experience delivering, scaling, and maintaining highly successful and innovative machine learning products.

Preferred Qualifications:

• Experience with Large Language Models: Demonstrated experience working with LLMs, such as GPT, BERT, or similar models, including knowledge of their strengths, limitations, and capabilities.
• Solid Understanding of NLP: In-depth knowledge of natural language processing (NLP) techniques and concepts, including tokenization, semantic analysis, and text generation.
• Have good understanding of state-of-the-art machine learning and deep learning technologies. In particular, hands-on experiences with deep learning models (DNN, Attention, CNN, RNN) and frameworks (TensorFlow, PyTorch, Keras, etc.) will be very helpful.
• Solid algorithm and analytical background and very good understanding on how to apply advanced knowledge to solve real problems.
• Ability to work independently in a team to deliver innovative solutions solving challenging business/technical problems from high level vision and architecture, down to quality design and implementation.
• Experience in parallel or distributed processing, high performance computing, stream computing and SCOPE.
• Self-motivated, self-directed, and be able to work constructively with a wide variety of people, team and changing business priorities.

Certain roles may be eligible for benefits and other compensation. Find additional benefits and pay information here:

• Building and maintaining production machine learning models for ad retrieval, quality prediction and creative generation.
• Finding insights and forming hypothesis on web-scale data with various machine learning, feature engineering, statistical, and data mining techniques: e.g. regression, classification, NLP, optimization, p-values analysis.
• Designing experiments, understanding the resulting data, and producing actionable, trustworthy conclusions from them.
• Craft and Optimize Prompts for Effective LLM Performance: Design, test, and refine prompts to elicit accurate, relevant, and useful responses from LLMs. This involves understanding the nuances of how the model interprets different inputs, experimenting with various prompt formulations, and iterating based on performance metrics and user feedback.
• Wrangling large amounts of data (think petabytes) using various tools, including open-source ones and your own. All programming languages are welcome, especially Python, R, C#, C++, Java, and SQL.
• Taking complex problems and the associated data and giving the answers in a concise form to assist senior executives in making key business decision.