Microsoft researchers reclaimed a third of their time for deeper analysis with Marvin

Senior researcher and anthropologist, Mary Gray, describes how she and her team cut down hours of manual work with Marvin.

Customer Outcomes

• Researchers save one-third of their time spent on manual tasks
• The team reinvests the time saved into deeper analysis
• Collaboration is significantly easier with a cloud-based, searchable repository
• Researchers easily keep participant data secure and compliant with privacy guidelines

The rigor of qualitative research meets the efficiency of AI

Anthropologist, Mary Gray, PhD, is a Senior Principal Researcher at Microsoft Research. Her work focuses on how people and communities engage with emerging technologies. Mary and her team rely on classic anthropological methods: field observations, in-depth interviews, and participatory engagement with communities.

Before Marvin, they were skeptical that an AI-centric tool would help them maintain qualitative rigor. But after consistently using it for in-depth studies, they discovered that Marvin facilitates research by taking on the time-consuming tasks of organizing and centralizing data. It does not aim to replace researchers, but instead makes them indispensable.

In some ways, we're doing deeper analysis than we could have done because of the constraints on time. That is a game-changer!

A cloud-first workflow built for easier collaboration

Mary works with interview transcripts, notes, photographs, and artifacts produced during fieldwork. Handwritten notes or visual artifacts that might previously have been lost as “unusable data” can be captured via photo, converted into text, and integrated directly into the analysis. Instead of spending countless hours manually sorting and coding the data, she can search for it in Marvin.

The team no longer has to print transcripts, highlight passages, and reconcile multiple versions across collaborators. With Marvin’s cloud-first design, distributed team members can collaborate in real time. Everyone can be involved in thematic coding sessions as they are no longer restricted by in-person meetings.

Mary uses Marvin to code, organize, and synthesize data more quickly. This frees up time for deeper analysis and writing. Rather than rushing through a coding pass, Mary and her team now revisit the data iteratively. They use the extra time to surface nuances and track subtle shifts in how participants describe their experiences over time.

I have never been able to collaborate so smoothly with colleagues who are not in the same place, in the same room. I'm easily saving about a third of my time with colleagues doing what we used to do, the hard way.

Trust is central to Mary’s work. Many of the participants she works with share highly sensitive information. Marvin’s secure, HIPAA-compliant architecture has become a key part of how Mary demonstrates responsibility and rigor.

Mary also uses Marvin to translate complex social science frameworks into accessible summaries for business stakeholders. She can effectively share her work without sacrificing nuance.

Once skeptical of AI tools, Mary now advocates for Marvin as a system that enhances, rather than replaces, human interpretation. By acting as a rigorous, collaborative sounding board, Marvin reinforces the irreplaceable role of the qualitative researcher. For Mary, Marvin has become an essential partner in conducting trustworthy, human-centered research.

To know if Marvin be an effective research partner for you, book time with us.