Nutrigap: user data you can ask questions of, with AI
The board and investors asked questions about the users that could not be answered. The answers were in the data, but compiling them by hand would have taken forever. Now Anna Hamilton asks the data herself.
- IndustryWomen's health and fertility
- EngagementAI and data analysis
- PartnershipProject
The problem
Nutrigap is an app in women's health and fertility. Anna Hamilton was getting questions from the board and investors about how users used the app and how that changed over time. All the data was there, but compiling it into reports by hand would have taken forever. So the reports never got made, and decisions were taken without them.
What we did
We gathered all the app's data in Google BigQuery and put an MCP server in front of it, so Claude can query the data directly. Anna writes her question in Claude Code and gets the answer compiled from the actual data, in the same conversation. No ticket to development, no dashboards built in advance for questions nobody has asked yet.
What it gave them
The data can now be asked, and the answer comes in whatever form the question calls for: a reply in the conversation, a table, a summary for the board as a PDF. Anna does not have to order anything from anyone, and the board and investors get numbers from real data instead of estimates. The next question costs no more than the last.