Microsoft Senior Applied Scientist 

United States, New York, New York 

621319784

Are you seeking opportunities at the intersection ofartificial inelligence (AI),such aslarge-scale multimodal models like GPT-4oApplied Scientist to join our team bridging responsible AI research and practice. The team focuses on sociotechnical alignment of generative AI systems, with a particular focus on measuring risks, including risk systematization, risk annotation, dataset creation, and metric design. As a member of this team, you’ll develop central resources and tooling, you’ll partner with Microsoft product teams who wish to be proactive about responsible AI, and you’ll contribute to and/or drive research projects intended to advance the state-of-the-art. Successful candidates will have experience working with large-scale language models and/or multimodal models, and will be passionate about prioritizing diversity, inclusion, and fairness.

Required Qualifications:

• Bachelor’s degree in computer science or a related field (e.g., statistics, information science) and 4+ years of experience working with AI, ML, and/or NLP. This can include product experience, industry research, open-source contributions, or academic research (but not coursework).

• Master’s degree in computer scienceor a related field and 3+ years of experience working with AI, ML, and/or NLP. This can include product experience, industry research, open-source contributions, or academic research (but not coursework).

• OR PhDin computer science or a related field and 1+ year(s) of experience working with AI, ML, and/or NLP. This can include product experience, industry research, open-source contributions, or academic research (but not coursework).

• OR equivalent experience.

• Experience putting responsible AI principles (e.g., fairness, transparency) into practice.

• Coding skills in a general-purpose coding language (e.g., Python, Java, Scala, C#, or C/C++).

Preferred Qualifications:

• Master’s degree in computer science or a related field (e.g., statistics, information science) and 6+ years of experience working with AI, ML, and/or NLP. This can include product experience, industry research, open-source contributions, or academic research (but not coursework).

• OR PhDin computer science or a related field and 3+ years of experience working with AI, ML, and/or NLP. This can include product experience, industry research, open-source contributions, or academic research (but not coursework).

• OR equivalent experience.

• 3+ years of experience contributing to or driving peer-reviewed academic papers.

• 1+ year(s) of experience presenting at conferences in the research or industry communities.

• 3+ years of experience conducting research as part of an academic or industry research program.

• year(s)of experience developing and deploying production systems, as part of a product team.

• Experience working at the intersection of generative AI and responsible AI, trust, and safety.

• Experience approachingAI risksas sociotechnical challengesrather thanpurelyalgorithmic ones.

• Experience working with products and services that incorporate generative AI.

• Experience working productively on an interdisciplinary team spanning multiple roles.

• Track recordof prioritizing diversity and inclusion in the workplace.

• be a member of an interdisciplinary team ofexperts onAIrisks.

• design and runexperiments that use API callsto generative AI systems.

• monitorresearch advances and draw on them to shape your applied science work.

• contribute to and/or drive research projects intended to advance the state-of-the-art.

• learn new skills and apply them as needed:e.g.,you might beasked tolearn about measurementmodeling andproduce a report on adataset’scontentvalidity.

• You’ll work with stakeholders with a variety of backgroundsin a variety of roles.

• present your work to internal and external stakeholders, including Microsoft leadership.

• develop central resources and tooling foridentifying, measuring,andmitigatingAIrisks

• work withpolicyand engineering teamstosystematizeAI risks.

• build andvalidateannotationguidelinesand datasetsformeasuringAI risks.

• work with Microsoft product teams togeneralize theseresources andmethodstobeappropriate fora variety of systems, use cases, and deployment contexts.