Canopy
Canopy is the part of DeepGarden people actually use. Someone opens it, types the question they would have put in an email to the data team, and gets an answer they can take into a meeting — the finding, the view that suits it, and the record of exactly where every number came from. There is nothing to configure and no query language to learn, because everything that makes the answer trustworthy happens underneath in Roots.
Ask
Plain-English questions, follow-ups that keep their context, and clarifying questions back when the request is ambiguous.
Views
The interface assembles around the answer: a trend becomes a chart, a ranking becomes a sortable table, a single fact becomes a single number.
Evidence
Every hard figure carries the query that produced it, the source it came from, the rows involved, and the moment it was pulled.
Ask
Type the question the way you would say it out loud. There is no syntax to learn and no request to file, and only the data you are cleared for is ever in scope, so the thing standing between you and an answer is the time it takes to type it.
Follow-ups keep their context. Narrowing to one region, one quarter or one product line is another sentence rather than another build, and when a question is genuinely ambiguous Canopy asks rather than guessing which of two readings you meant.
Views
A trend arrives as a chart. A ranking arrives as a table you can sort. A comparison arrives side by side. A single fact arrives as a single number with its evidence attached. You do not choose a visualization or configure anything, because the shape of the answer follows from the shape of the question.
Every dashboard an organization owns is a standing liability: definitions drift, sources change, panels quietly break, and somebody has to keep the whole estate honest. Generated views have nothing to rot because nothing is stored waiting to. They are produced fresh from live data and current definitions each time they are asked for.
Anything worth returning to can be pinned into a shared workspace for the people who need it, with the same citations attached and the same live refresh underneath. The difference is that these views come from questions people actually asked rather than from a specification written in advance.
Evidence
A good analyst never hands you a number on its own. They hand you the number, where it came from, what was counted, what was left out, and when they pulled it. That is chain of custody, and it is the reason you trust their work in a room full of people who will push back. Canopy attaches the same thing to every figure it produces, automatically.
Click any hard number and it opens its own record: the exact query that produced it, the dataset it was drawn from, how many rows were involved, and the timestamp of the pull. Nothing is reconstructed after the fact — the record is created in the same run as the number, so a figure in a deck from three weeks ago can still tell you what it was standing on.
Evidence is not a premium feature and never will be. There is no cheaper configuration of DeepGarden that removes traceability, because an answer you cannot check is not a cheaper answer, it is a different product.
Why it changes the meeting
Most analytics disputes are not really about strategy. They are about provenance. Two people bring two versions of the same metric, nobody can settle it in the room, and the decision slides a week while somebody goes and checks.
When every figure carries its own evidence, that detour disappears. The question in the room becomes what to do about the number rather than whether to believe it.
Canopy by DeepGarden — governed natural-language analytics, for everyone you are willing to give access to.