The company brain
What actually moves the cost of a company brain
Last updated
Short answer
There is no credible published rate for this work, and any figure you find is somebody quoting their own price list as though it were market data. Five things actually move it: the condition of the material you already have, how many systems it must connect to, how many jobs it does before you judge it, how much regulatory exposure sits in the process, and who keeps it current afterwards. The fifth is the one most quotes leave out and the one that decides whether the thing is still working in a year.
Start here
Why there is no market rate to quote
Search for what a system like this costs and you will find confident numbers. Check where they come from and almost all of them are published by companies selling the service, quoting their own commercial terms as though they were an industry benchmark. There is no survey behind them, and the category is too new and too varied for one to mean much yet.
So rather than repeat figures that cannot be stood behind, here is what can be said honestly about what moves the number. Treat all of it as supplier material, including this page.
The framing that helps most buyers is not comparing quotes against each other.
The question is whether the work this removes is worth more than the work of building it. If you cannot describe that work in one sentence with a number attached, the quote is not the problem yet.
The drivers
The five things that actually move it
- 01
The condition of your material
Not how much you have, how consistent it is. A well-organised set of current procedures is inexpensive to build on. Three conflicting versions of the same policy across two drives and an inbox is the single largest and least predictable driver, and it is usually invisible until somebody looks.
- 02
How many systems it connects to
Each connection is a piece of work with its own quirks, permissions and failure modes. Two is very different from nine, and the ninth is rarely the one anybody remembers to mention at the start.
- 03
How many jobs it does before you judge it
One bounded job is a different engagement from a shared system covering four departments. Starting narrow is not a discount, it is how you find out whether the material is in good enough shape to build on before committing to the wide version.
- 04
The regulatory exposure in the process
A process touching personal data, money movement or a regulated activity carries review, documentation and oversight obligations that an internal knowledge lookup does not. This is real work rather than paperwork, and it is not optional.
- 05
Who keeps it current afterwards
A system nobody maintains degrades quietly, which is worse than one that fails loudly. Whether that is you, us, or a shared arrangement changes both the ongoing arrangement and how much has to go into onboarding your people up front.
Avoidable
What makes it more expensive later
Three of these are avoidable at no cost by deciding differently at the start.
- Starting wide instead of narrowConnecting everything before knowing whether the first job works. When it underdelivers, and something usually does at first, you cannot tell which of the nine connections was the problem.
- Skipping the written blueprintThe most expensive mistake in this work is building the wrong thing competently. A document agreed before the build is the cheapest insurance available against it, and it is the deliverable people are most tempted to skip.
- Leaving your people out of the onboardingIf nobody inside the company can teach it a new job, every future job comes back to the supplier, forever. That is a recurring line on your side and a dependency you chose without noticing.
- Fixing the material during the build rather than beforeSorting out conflicting procedures is work that has to happen either way. Doing it inside a build means paying build rates for document archaeology and discovering scope as you go.
Judging a quote
What you should refuse to pay for
These come up often enough to be worth naming, in any quote from anyone.
- A per-person charge for a system that answers questionsIt is priced that way because that is how the software industry prices seats, not because a second person asking costs materially more. It also creates exactly the wrong incentive: you limit who can ask, and a system half the company cannot reach does not accumulate anything.
- A build with no named first jobIf the proposal describes capabilities rather than one job somebody does weekly, nobody has yet decided what success looks like. That decision does not become easier after the money is spent.
- Anything described as a pilot with no defined endA pilot with no agreed number that counts as working becomes a permanent pilot, and permanent pilots are how budgets are consumed without a decision ever being made.
- A premium tier justified by making the standard one sound unlawfulStricter hosting arrangements buy supply-chain control, which some companies genuinely need. Anyone implying their normal product is not compliant is telling you something about their normal product.
- Ongoing charges with nothing ongoing attachedA recurring arrangement should buy something recurring: the material kept current, jobs added and retired, and somebody reviewing what people ask and where the answers are wrong. If it buys availability alone, ask what happens in a month when you need nothing.
Before you compare quotes
The number that matters more than the quote
Work this out before speaking to anyone, including me, because it makes every subsequent conversation shorter and it is the one piece of arithmetic no supplier can do for you.
Take one job. How many times a week does it happen, how long does it take, and who does it. That gives you what one occurrence costs today. Then ask what proportion of it is genuinely repetitive rather than judgement. The repetitive part is the only part any system of this kind removes, and quotes should be judged against that number rather than against each other.
Frequently asked questions
Why will you not publish a figure?
Because scope rather than seniority sets it, and a published figure would be either meaningless or misleading in most situations it is read in. Every engagement starts with a fixed-scope piece of work that produces a written plan, so both sides know what the work actually is before anyone commits to it.
Is it cheaper to start small?
It is cheaper to start narrow, which is not quite the same thing. One bounded job with a real deadline and a named owner costs less and tells you more than a broad system built on assumptions. The saving is in what you avoid building rather than in a discount.
What is the ongoing arrangement actually for?
Keeping the material current, adding and retiring jobs, and reviewing what people ask and where the answers are wrong. Those are the activities that decide whether it still works in a year, and they are the ones most likely to be quietly absent from a quote that looks competitive.
Can we build this in-house instead?
Some companies should, particularly where there is an engineering team with capacity and the material is already in good order. The part that is consistently underestimated is not the build, it is the maintenance and the access design, which is also where in-house attempts most often stall.
How do we compare two quotes that look different?
Put both against the same named first job, ask each what they are not connecting and why, and ask each what the recurring arrangement buys. Quotes that look far apart usually differ in scope rather than in rate, and those three questions surface it.
AINORA, MB builds and operates AI systems for companies from Vilnius, Lithuania. It designs company brains: shared internal systems that hold an organisation's own knowledge, connect to the tools it already runs, and carry out the jobs its people repeat, with access rights enforced per person and hosting in the EU.
Bring one job and its arithmetic
A conversation about a specific job, how often it happens and who does it is worth more than any comparison of proposals. If the arithmetic says it is not worth building, that is what you will hear.