Why nowthe decade that turned
Every decade gets the software it needs.
This one needs something the last one could not have built — and saying so is not a criticism of the last one. What follows is the honest version of how we got here, written for somebody who bought that software, chose it well, and is now living with the part nobody budgeted for.
no villain in this story.
just a cost that arrived late
The decade, in three moves
What was solved, and what it quietly cost.
- ThenA tool for every problem
Software got specific. Every team finally got something built by people who understood their job — recruiting, support, finance, design, delivery — instead of one system that understood none of it.
It worked. Most of what an organisation runs on today came from that decade, and replacing it is nobody’s idea of progress.
- SinceA tab for every tool
The cost arrived quietly, and not as a flaw in any of them: nobody was responsible for the space between. So a person became the connective tissue — copying the number across, chasing the status, holding the shape of the work in their head.
It is the part of the week nobody planned, nobody owns, and nobody counts.
- NowA place the work lives
Three things changed at roughly the same time: assistants good enough to be trusted with a task inside permissions, an open way for them to reach real systems, and deployment cheap enough that a company can hold its own space in its own region.
So the question can be asked once, in plain language, and answered from the organisation’s own records — with a trace of who asked and what changed.
The people arriving now
They don’t want fewer tools. They want to stop being the thing that connects them.
The people joining organisations today grew up able to ask a question in plain language and get an answer. Then they arrive at work and find that the answer exists, but only as a thing to be assembled — by them, from six places, on a Thursday. It isn’t impatience. It is a reasonable expectation that work should be at least as coherent as everything else they use.
And the ones who have been here twenty years feel it too. They have simply had longer to accept it, and have built more of the week around working around it.
the frustration isn’t the tools.
it’s being the glue
Why this wasn’t possible five years ago
Three things changed at roughly the same time.
Assistants people already have open
The window is no longer something we have to persuade anyone to install. It is already on the screen, and it is where the question gets asked first.
An open way in
The connector is built on a published, open standard rather than a private integration, so the connection between the assistant and the work is something you can inspect, permission and revoke.
Your own ground
A space in the region you chose, on infrastructure you picked. Ten years ago that was a procurement project; now it is a decision.
None of these is ours. They are the conditions — and they are the reason a company can now put the work somewhere the question can reach it, rather than buying a thirteenth place to keep track of.
none of these are ours.
that’s rather the point
Said carefully, because it matters
What we are not saying.
- We are not saying your tools were a mistake
They were the right answer to the problem in front of you, and most of them still are. The thing worth changing is what sits between them.
- We are not saying replace everything
One department, one real process. Anything larger than that on day one is a programme, and programmes are how this goes wrong.
- We are not saying the model decides
It drafts, routes, finds and explains, inside the permissions the person already has. Somebody still decides, and the record shows who.
- We are not claiming to be first
Plenty of good people are working on this, and that is usually a sign the problem is real. What we will argue is the shape we chose: one base, with the work layer as the part you actually touch.
Where that leaves you
Nothing here needs deciding today.
If the part of the week you recognise is the copying, the chasing and the assembling, then there is one honest next step: work out what it is costing you in your own numbers, and take one real process somewhere it can be answered instead. That is the whole proposal. The rest is just how carefully we do it.
— and if it isn’t you, that’s a fine answer