The invoice inbox
Invoices arrive as email attachments. Somebody opens each one, keys the coding, matches it against the purchase order, and chases what does not match.
We take the manual work your people do between systems and handle it end to end. You own the code. The return is measured before anything is written.
Each one is a person, in a chair, moving something from one system into another because the two will not do it themselves. They are the first thing we count, because they are where the money is.
Invoices arrive as email attachments. Somebody opens each one, keys the coding, matches it against the purchase order, and chases what does not match.
The same workbook gets rebuilt every period. Reconciliation against the source systems, variance notes written from memory, and a close that lands late.
It arrives as a PDF, in the body of an email, or on a form nobody controls. Somebody reads it, finds the customer, checks the pricing, types it in again.
One person holds the sequence, the constraints and who is where. When something moves they rebuild the day in their head, then tell everyone twice.
More requests than hours. Triage by hand, access and onboarding runs typed one at a time, and the same three answers written out again this week.
Certificates, contract dates and filings sit in a folder, in a calendar and in one head. The first sign that one lapsed is usually the consequence.
Someone will ask what this returned. Most firms cannot answer, because nobody measured the work before they changed it. So we measure it first: what one invoice, one ticket, one job costs you today. Then we build. Then we measure it again and publish the result whether or not it flatters us.
The number is taken in the Scoping Week, before anything is priced.
We build one run at a time, end to end. Not the task, the whole run: from the attachment landing in the inbox to the record being correct in the system your team already opens. The Scoping Week decides which run, and what it costs.
$5,000
one time, five days
≈ a decision, not a proposal
You bring the runs your team has been circling. We put each one through four questions, model the return from your own figures rather than an industry average, and come back with the one worth building, priced, with an owner named. It can end with a recommendation to build nothing, and that has to be a real possible output or the week is theatre.
Not in: the build itself. The Scoping Week buys the decision and the price of the build, not the build. Also not in: no code, no pilot, no maturity score out of five, no board deck.
Fixed fee, five days. Credited in full against the build.
Fixed
six weeks, priced in the Scoping Week
≈ a system you own, not a demo
Six weeks against a written scope, at the price agreed before it starts and unchanged once it has. Built inside the systems your team already opens rather than alongside them, with a working demonstration every week so nothing arrives as a surprise at the end.
Not in: standing access to production data after handover unless Care is signed, model hosting on our account, and any work where there is no way to tell afterwards whether it worked.
The price is agreed before the build starts and does not move once it has.
$2,500 to $5,000
per month, cancel any quarter
≈ 18 to 20% of what the build cost, per year
One or two live systems, watched. You pay whichever is greater: $2,500 a month, or 18 to 20% of what the build cost, spread across the year. On a smaller build that means the $2,500, because watching a live system takes the same work either way.
Not in: no new builds and no roadmap work. Anything with a scope on it is a build or an Embedded month, not Care.
Cancel any quarter, on the report rather than on an anniversary.
If what you want is continuous shipping rather than a build with an end date, Embedded runs $8,000 to $15,000 a month on ninety days notice, re-scoped every year against what was actually measured. It is shipping capacity rather than maintenance, so it is not priced as a share of a build, and nothing here runs longer than twelve months without a renewal decision.
What you own at the end is not a report on what we did. It is the working system and everything needed to run it without us. Every item below is on the price list, which is why it can be printed here rather than discussed on a call.
Runbooks written for the person who has to restart it at seven in the morning, rather than for the person who approved it.
It runs on infrastructure you own. Standing access to production data after handover is not included unless Care or Embedded is signed, and that exclusion is on the price list rather than in the small print.
The model is locked so it cannot change under you without warning, and you keep a set of tests that catches it early if the answers start to drift.
Model vendors publish how much notice they give before retiring a model. Watching that schedule against what you are running is part of what Care does.
No six month onboarding and no drawn out process. One signature gets you the Scoping Week. What it finds decides whether there is a build at all, and it is priced before you commit to it.
One fixed fee of $5,000, under most approval thresholds. Its whole cost comes off the build if there is one.
Time with the people doing the work. For each run we capture who does it, how long it takes, how often, what it runs on, and what happens when it goes wrong. That last question shapes the design more than any of the others.
Each run put through four questions: can it be built here, is the arithmetic real, how you will know it worked, and what happens when it is wrong. One recommendation, priced, with the reasoning for choosing it over the rest.
Build with us, build with somebody else, or do nothing. If it is a build, six weeks against the written scope at the price already agreed, with a demonstration every week.
That list is six shapes, not a catalogue. What matters is not whether your run is on it but whether it looks like those: a person moving something between two systems, often enough to count, where somebody could tell you today roughly what it costs to do by hand. If that is your run, bring it. If you have six of them and no way to rank them, that is what the Scoping Week is for.
Yes. The Scoping Week is $5,000, flat, and its whole fee is credited against the build if there is one. There is no call standing between you and that price, which is the entire point of publishing it. What it buys is the decision: the runs in your company ranked, one recommendation, a verdict on whether it is buildable, and the price of building it, in writing. The build itself is priced separately, once we know what it is.
Because the Scoping Week already did the hard part: choosing the one run worth building and confirming it can be built. Six weeks is what is left once that decision is made and the scope is fixed. It runs inside the systems your team already opens rather than alongside them, with a working demonstration every week, so nothing arrives as a surprise at the end.
Then the Scoping Week says so in writing, before a price is agreed rather than after. If the right first project turns out to be a data problem rather than a build, that is what the recommendation will say. It is a cheaper way to find that out than a build that discovers it in month two.
If they have six weeks free, they should. What usually stops this work is not that nobody knows how. It is that the people who know how are the people already holding the operation together, and a build running alongside daily work is the one that slips. If your team has the room, that is the cheaper answer and we will say so.
The Scoping Week records who touches the run today, on every run you raise. The recommendation names which steps come off a person and which do not, before anything is built, and the arithmetic is modeled from your own figures rather than an industry average. The person whose team would have to check the output is interviewed during the week, not at kickoff of a build.
You do. The code, the documentation, the runbooks and client-owned infrastructure are part of the handover and they are in the price rather than an extra line on it. Standing access to production data after handover is not included unless Care or Embedded is signed.
Name the run that costs you most. Forty five minutes, no deck and no discovery theatre, and you leave knowing whether it can be built, roughly what that takes, and whether it is worth doing at all.