A fair questiontaken seriously
Yes. It's crowded.
We get asked how this differs from the platforms that already exist, and whether there is a moat. Both are fair, and the honest answer starts by agreeing with the part that is true.
If this were another work-management tool, there would be no argument.
The category is well served. There are excellent, well-funded products in it, and they got there first. Building a slightly better board would be a poor use of everyone's time — so that isn't the bet.
What is actually being asked
Not which is better. Which market this is.
- A work tool with AI addedthe crowded one
Well served, and nothing here changes that. If that is what you need, the established products are a sound choice and we would say so.
- Running work without opening onethe bet
The ask happens in the assistant you already have open, through a connector, and the status is read from the work rather than typed by someone on a Friday. That is not a better board — it is a different question.
The honest test: if this ends up as a good board with a chat panel beside it, the objection was right and we should say so first.
Three structural choices
Not features. Features get copied.
Adding, or removing
The common move is to place an assistant beside the workspace — a panel, a sidebar. It works well, and it keeps the screen at the centre. Our bet removes the screen instead. Two different bets, and a company that sells seats which open a dashboard has good reasons to prefer the first.
Record, or context
Established platforms hold the system of record: what happened, stored well. Our bet is that the value moves to context — history, permissions, semantics and consequence, connected. A record exports cleanly. A context graph is built by being used.
Your account, your keys
Multi-tenant software cannot offer "runs inside your own cloud" as a pricing tier; it is a different shape of company. For regulated and sovereignty-minded buyers that is not a preference — it decides the whole thing.
On moats
Moats come later. Wedges come first.
A moat is the result of having won something — switching costs, accumulated data, distribution. It describes where a company ended up, not the plan that took it there. Every moat worth having was a wedge first, and the wedge is the part a company actually chooses.
So the question that predicts anything is not what is your moat. It is what compounds as this works.
the honest bit — worth
reading twice
- Contextcompounds
The graph of decisions, permissions and history an organisation accumulates. Rebuilt only by re-living the same months of work.
- Deploymentcompounds
Spaces running inside customers' own infrastructure. Moving out of your own cloud is a project, not a cancelled subscription.
- The ecosystemcompounds
Nine layers shared across verticals, so each new one costs less than the last. A single-product company re-pays full price every time.
- The connectornot a moat
An open standard, deliberately — that is the point of it, and anyone can implement it. Saying so is what makes the three above worth believing.
Testable
What would change our mind.
Positioning that cannot be tested is marketing. Any one of these inside twelve months is a signal to change the bet rather than the copy.
- Assistants read status from the work01
If the established suites ship assistants that write to a plan and report its state without anyone opening it, our surface argument is gone.
- Nobody asks in the assistant02
If people add the connector and keep opening the web app anyway, the core assumption is wrong and this should become a very good work tool instead.
- A search box answers as well03
If retrieval over a system of record matches a context graph, the data-model argument was a story.
Fit
Who this is for.
- Happy with your current toolnot a fit
Genuinely — nothing here is worth a migration, and we would say that on a call.
- No sovereignty requirementnot a fit
If your suite already covers the work and nobody asks where the data sits, it is the sensible answer.
- Want AI added to how you worknot a fit
That is AI-enabled, the market for it is well served, and it is not this.
- Admin became most of the worka fit
And somebody senior cares which country the data sits in.
A product that is for everyone is for nobody.
Tell us where we're wrong