Amazon Web Services operates at a scale few technology companies can match. Thousands of engineers, designers, and product managers build across hundreds of services—each team making rapid decisions about what to build next, how to improve onboarding, and where customers are running into friction. Beneath all of that sits a research practice that spans the full breadth of the organization: UX researchers running moderated and unmoderated studies, product teams synthesizing open-ended survey responses, customer experience functions ingesting support signals, and design teams trying to stay close to users while shipping at pace.

The challenge wasn't a lack of customer data. It was that the data was everywhere. Research documentation lived "in all four corners of the world," as one researcher put it—interview recordings stored in Chime, synthesis canvases built in FigJam, UserTesting exports living in shared drives, and insights that existed only in the memory of whoever had run the study. When a new team member joined, or when a product manager needed to know whether a particular customer problem had ever been explored, there was no reliable place to look.

AWS didn't come to Dovetail through a single champion. The relationship grew team by team—a researcher here, a design lead there—until Dovetail became embedded across functions including Books Payments, the People Experience team, Amazon Health, Whole Foods International Seller Services, the ADX team, and multiple AWS developer experience groups.

What made that expansion possible was a pre-negotiated enterprise agreement with Amazon legal, giving any team a fast path to getting started. The commercial structure was designed to remove friction: a flat platform fee, unlimited data storage, unlimited AI queries, and unlimited viewers so stakeholders across a team can access research without requiring a paid seat. Contributors—the researchers and designers doing the synthesis work—are the only licensed role, keeping cost predictable as teams grow.

The result is that a researcher at Amazon Leo, can get started on the same terms as a team at AWS that has been using Dovetail for years. When a new team sees a demo and decides to move forward, the path from decision to active workspace is measured in days, not quarters.

For the teams at AWS managing high volumes of customer feedback—survey responses, app store reviews, support transcripts, open-ended study data—the analytical challenge is consistent: how do you find the signal in thousands of unstructured text responses without spending weeks doing it manually?

Dovetail Channels addresses this directly. Teams pipe their feedback sources into a Channel, define the topics they care about, and Dovetail's AI clusters the data into themes, tracks how those themes trend over time, and generates summaries grounded in the actual language customers are using. When a product manager wants to know whether a specific complaint is growing or shrinking after a recent release, they can see the volume curve and drill into the quotes behind it.

Contextual Chat takes the analysis further. Because Chat is scoped to whatever context a user is in—a single Channel, a folder, or the whole workspace—teams can ask natural language questions against their specific data set and get answers with citations back to the source. That citation chain matters at AWS: when a researcher presents a finding to an engineering team or a product leader, the ability to trace any claim back to a verbatim customer quote builds the kind of confidence that drives decisions rather than further debate.

One researcher described the goal simply: analyze across multiple studies over time, synthesize the pattern, and share it with the people who need it before they go off and commission the same research again.

The breadth of the AWS and Amazon partnership with Dovetail reflects something larger than any individual team's research workflow. When more than 25 teams in a single organization converge on one platform, the value stops being about any one team's efficiency and starts being about the organization's collective intelligence.

A researcher joining a new team can search Dovetail and discover what customers said two years ago about a problem they're about to study. A product manager can pull a thread from a UX study and connect it to themes emerging in support tickets. A design leader can share a link instead of a deck, and the stakeholder on the other end can go deeper on their own—reading quotes, replying to insights, and building a direct relationship with the customer data rather than consuming a filtered summary.

The infrastructure that powers Dovetail's AI is built on AWS Bedrock, which means the two organizations' relationship runs deeper than a software subscription. Dovetail's position as a customer of AWS—and AWS as a customer of Dovetail—reflects a shared belief that the distance between customer feedback and product decisions should be as short as possible. At AWS, the work of closing that gap is already underway, one team at a time.