Context
UpdateAI helped account and customer success teams turn customer conversations into meeting, account and portfolio insights. The company was later acquired by Gainsight to strengthen its Atlas agentic AI strategy.
The problem
The company had done significant research, but it had not become product decisions. Three questions were open: what the MVP should be, which problem it should solve first, and how to validate the experience before building it.
What I did
I led a one-month design sprint, adapted for early-stage product definition. The standard sprint format of strategy, design, prototyping and user testing was stretched over a month to leave room for research synthesis, iteration and MVP planning.
I also facilitated the workshops and helped the client make decisions based on the test results.
How I worked
- Synthesized scattered research into a focused product direction, moving from broad opportunity areas to a concrete experience with clear priorities.
- Defined the MVP scope. We used the user tests to learn what mattered most to users and what had to stay, which let us cut the scope.
- Built a prototype and ran the first round of user tests with five participants over three days.
- Changed the design based on what we saw, then tested again.
Key decisions
Build on the research they already had.
We analyzed all of it, researched users ourselves, and shaped an MVP that made sense for what their target audience wanted.
Outcome
- A validated product direction before development started.
- A full MVP development roadmap.
- A finalized brand direction and style guide.
What I took from it
A team can have plenty of research and still not know what to build. What helped was giving it structure: a sprint, real users testing a prototype, and a clear view of what mattered most to them. Facilitating the workshops showed me that part of my job is helping a client decide, not only designing the options.