The most valuable thing your organization owns isn’t on any list of things it owns.

Every organization owns knowledge it can’t find, can’t keep, and can’t hand over.

You can produce a list of your buildings. A list of your machines. A list of your data, your patents, your contracts. Somebody signs for those. Somebody notices when one goes missing.

Now try to produce a list of what your organization actually knows.

Not what’s in the manuals — what’s in the people. The one who knows why the line does that in humid weather. The one who can hear a bearing going. The one everybody calls before they call the vendor. The one who knows which step in the protocol the printed version gets wrong.

None of that is on any ledger. It has no owner, no custodian, no chain of custody. It is, by some distance, the most valuable thing in the building — and it is the only asset that can leave on a Friday without anyone booking the loss.

Know-how is the one asset with no entry in the books.

For most of history, this solved itself.

Nobody worried about it, because expertise moved hand to hand. You stood next to someone who knew, for long enough, and eventually you knew too. Nothing was written down because nothing needed to be. The transfer happened in the doing — in questions asked over a machine, in being corrected, in watching someone decide.

That mechanism is under pressure from two directions at once.

The people carrying the knowledge are leaving faster than the people who could learn it from them are arriving. That much is widely understood, and it is usually where the conversation stops.

The second pressure is less discussed and more serious. The junior work where judgement used to be built — the slow, repetitive, unglamorous reps that nobody enjoyed and everybody needed — is steadily being automated away. That work was never only production. It was the apprenticeship. Remove it and you have not just lost some tasks; you have cut the line along which expertise used to travel.

The first problem is an inventory problem, and you can feel like you’re solving it by hiring. The second one you cannot hire your way out of, because the people you would hire are the ones who were supposed to be learning this now.

The experts are leaving. The machine that made new ones is stalling.

Organizations have been trying to fix this for forty years. Mostly by writing it down.

The attempt is honest and it keeps failing, for reasons that have nothing to do with how hard anyone tried.

Writing it down is the expert’s least favourite afternoon. So the knowledge you most want captured belongs to the person least willing to sit down and type it, and it gets documented last, or never.

A document is finished the day it’s written and wrong six months later. The process changed, the machine was replaced, somebody found a better way — and nobody was assigned to notice.

A document answers the question someone thought to ask in advance. Work produces the other kind: the question nobody anticipated, at the moment the line is stopped and three people are standing there.

And storage systems are organized around where things live, not around the moment someone is stuck. Knowing that the answer is in a folder somewhere is not the same as having it.

The industry built for this problem — course-based training — made it worse in one specific way. It inverted the sequence. Learn first, then work. Real work does not wait behind a curriculum, and the moment of need does not arrive on a schedule.

Storage is not retrieval. A folder is not an answer.

So we built the instrument we thought was missing.

One requirement above all others: it had to work without asking experts to become writers. If capture costs an afternoon, the best knowledge never gets captured, and every system that ignores this fails in the same place.

So an expert shows us once. Voice, video, a screen recording, or a template — whatever is least in the way. No script, no storyboard, no instructional designer, no edit pass. What comes out is a guide: structured, mapped to the machine or process or role it belongs to, kept current, and able to answer a question in the moment somebody is stuck.

Then we ran into the thing that made this two products.

Knowledge sits under two opposite pressures, and most organizations have both. Some of what you know is precisely the thing a competitor would pay for. It should never leave your walls — private, governed, inside your boundary, under your access rules. And some of what you know is worthless until it reaches people outside those walls: your members, your clients, the people using your product who need to know how it actually works.

Same capture. Opposite intent. Trying to serve both with one product means one of them is always a compromise — a private system with sharing bolted on, or a publishing tool pretending to be secure.

So DeemL Core activates knowledge inward, for your own people. DeemL Hub activates it outward, for your audience. One engine underneath, two directions of travel.

Some knowledge should never leave your walls. Some is worthless until it does.

One thing we’re careful about.

It would be easy to describe all of this as a way to get knowledge out of people before they go. That framing is common in our industry and we think it’s both unpleasant and wrong.

The people who hold this knowledge are not a risk to be managed. They are the reason the place runs. What is actually true is stranger: their judgement is invisible to the organization that depends on it most. Nobody can see it, so nobody can credit it, and it gets noticed for the first time on the day it’s gone.

Making it visible is not extraction. It is the first time the organization can see what it has been relying on — and the first time the person who built that judgement gets credited for something more durable than being the one everybody calls.

Their judgement is invisible to the organization that depends on it most.

Why this matters more the longer you wait.

Captured know-how compounds. Every procedure captured makes the next question answerable, and the one after that easier to ask. A body of knowledge two years old is not twice as useful as one a year old — it’s the difference between a system people check and a system people trust.

Which means the gap between an organization that starts now and one that starts in three years isn’t three years of effort. It’s three years of sessions that can’t be run retroactively, with some of the people who could have given them no longer there.

That’s the honest version of urgency. Not that something terrible happens if you wait — but that some of what you’d be capturing has an expiry date attached to a person.