The forecast finally has a paper trail
Maya — Director of FP&AMid-size consumer products company
Maya bought a Team plan for her group. The CFO was both her champion and the person who signed for it. Because finance controls its own tooling spend, there was no IT budget owner to convince and no procurement queue. She decided on a Tuesday and the team was working in it that week.
What tipped them over
What tipped Maya was a near-miss. An analyst left mid-cycle, and a driver in the revenue model turned out to live only in that person's head and a tab nobody else understood. The number held, but barely, and the reconstruction ate two days she didn't have. She wasn't looking for another dashboard. She wanted the context around the work to stop walking out the door. This was a deliberate contrast to how tooling usually got bought at the company: a business owner making the case while a separate IT team held the budget. Here the sponsor and the budget were the same office, so it moved.
The first problem they solved
The team started by cataloging what they actually owned, the revenue build, the headcount plan, the cash forecast, the board reconciliation, as metadata: names, owners, sources, and what fed what. No values left their files; MyDataWork only ever held the map, not the money. Then each model got framed as a use case with its stakeholders and its worth. It fit the way they already worked. Analysts kept their spreadsheets and queries; they just described them once, and the in-app AI drafted the plain-language summary of why each one mattered.
"For the first time I could show the CFO the whole portfolio, what it's worth, what it leans on, without opening anyone's working files."
Scaling from here
The change the CFO noticed came from the aggregates-only Architecture and Value rollup. In one admin view Maya could show what forecasting work existed, what each piece was worth, and what it depended on, without exposing a single analyst's private draft. When the board asked where a growth assumption came from, she traced it to its source in minutes instead of a scramble of emails. Board prep stopped being a fire drill and started being a review of something that was already documented.
Toward agentic & generative AI
Maya isn't rushing to build a finance AI assistant, but she wants to be ready when she does. Standardizing the team's work estate is the groundwork. In Agent Studio she's already sketched which close and forecast-prep steps could be handed to an agent, scoping the task before committing to build. And governed Agent Access means any assistant she eventually approves reads only the metadata context, the shape of the work, never the underlying figures.
