How it works
You pick one workflow. I work it out end to end. You get a document you keep, whatever you decide next.
Most companies your size have someone who keeps the computers running, or a supplier who does. What almost nobody has is someone whose job is to make the work itself smaller. That is the gap the audit fills, and it starts with one process, not with your whole company.
What the audit is
You choose one workflow that feels heavier than it should: the thing three people touch before it is done, the numbers that get typed from one system into another, the request that sits in an inbox until someone has time. I take that one process and work it out, from the first step to the last, in about a week.
It is not a scan of your entire company, and it is not a strategy deck. It is one workflow, on paper, with a number next to it.
The document you get
At the end you have a short document. Six parts, no filler:
- The summary. Your situation in plain words, what I propose, and the expected outcome. One page, readable by someone who was not in the room.
- How it runs now. The actual steps, who does each one, and where the work waits or gets re-entered by hand.
- What it costs you now. In hours per week and euros per month, using your numbers, not mine. This is the part that tells you whether it is worth fixing at all.
- What I propose. The fix, concretely: which tools, what connects to what, and what changes for the people doing the work. Often this is not AI.
- The plan. In phases, each with something working you can see, so you are never asked to trust a promise you cannot check.
- What it returns. Hours and euros saved, and how many months until it has paid for itself. Deliberately on the careful side.
You keep this document whether or not you hire me to build anything. If the honest answer is that the workflow is fine as it is, that is in there too, and you have still learned something worth knowing.
If you decide to build it
Then the document becomes the plan, and the build runs the way I have always worked:
- Fixed price, agreed before I start. Not a ten-page quote, a few lines we both understand. If it runs over, that is my problem, not your bill.
- Built in phases, each with a check. After every phase it is settled whether it is right. Not "it should work", but demonstrably: the number matches, the page loads, the mail arrives. A phase that does not check out does not move to the next.
- Verified working before anything goes live. The new thing runs and is checked while the old thing still works. Only when it demonstrably does what it should does the switch flip. And you pick that moment, not me.
This is not a method I invented for this page.
This summer I moved two practices off shared hosting onto their own server, where I run them now. The plan was on paper first: every phase with its own check, and a hard stop just before the irreversible moment. Both sites ran completely and were verified on the new server before anything at all was switched over. The owners chose the moment themselves.
That is not modesty. It is how you stop a migration from costing an evening instead of a minute.
And after: I watch it without you noticing
Anything I run has monitoring on it. Every minute a check: is the site still up, is the mail still flowing, is there room on the disk. If something breaks I get a message on my phone, usually before you notice.
For my oldest measured client this has run that way for years. Over the last twelve months their whole environment was reachable 99.86% of the time. That is not a promise from a brochure, it is what the monitoring says, and I can show it to you.