Stage one: an AI readiness assessment
Find out what is worth automating before anyone builds anything.
Two weeks. I map how your work actually moves, mark what is safe to automate and what has to stay with a person, check it against the compliance you already have, and price the build in phases.
The AI readiness assessment: four questions, answered in writing
Where is the manual work really?
Not where you assume it is. I follow a real piece of work end to end and record every point where a person retypes, re-checks, chases, or re-decides something. Rough hours attached, so the expensive steps become obvious.
What is safe to automate?
Each step marked: automate fully, automate with a human checkpoint, or leave alone. The third category matters most. Anything involving professional judgment, a duty of care, or an irreversible consequence stays with a person, and I will say so plainly.
What does your compliance already permit?
What your existing agreements already cover, where data would move, and whether anything new would need signing. Frequently the answer is that you can build inside what you already have, which removes the slowest part of the project.
What would it cost, and in what order?
A phased plan with a price against each phase, ordered so the earliest work returns the most. You should be able to stop after any phase and still be better off than when you started.
Want a rough number before we talk? The manual work cost calculator takes three of your own figures, with no email required.
How the two weeks run
Kickoff, about an hour
You walk me through how the work comes in and what happens to it. I mostly ask questions and try not to propose anything yet.
I follow the work
Reviewing how things actually move, where they stall, and what your systems already do that nobody is using. Short follow-up questions as they come up, not a standing meeting.
Compliance and feasibility
What your existing agreements cover, what the relevant rules require, and what is genuinely buildable against your current stack rather than in principle.
The plan gets written
Workflow map, automate-or-not marking, phased build plan with prices. Written to be read by you, not by a developer.
Walkthrough, about an hour
I take you through it, you push back, I adjust. Then it is yours, whatever you decide to do next.
This fits if
- +You run a firm, practice, or operation where a wrong answer has real consequences
- +Work arrives faster than anyone can triage it properly
- +Your systems are fine but the gaps between them are manual
- +You have looked at AI tools and stalled on whether they can be trusted
- +You want to own what gets built rather than rent it
This does not fit if
- −You want someone to run a chatbot pilot and call it a transformation
- −The real problem is that a process is undefined, not that it is manual
- −You need a decision this week, since the honest version takes two
- −You want AI making the professional judgment call, which I will not build
Further reading
AI readiness assessment: a test for what not to automate
an AI readiness assessment methodology that tells you which decisions the model should never make, not just what to buy. a test you can run yourself.
what business processes can be automated (and what can't)
what business processes can be automated? the test isn't repetitive vs. rare, it's fact vs. judgment. six steps to find the real boundary.
how to run AI on client data without a compliance headache
a practical way to put AI on client data without handing it decision power: split reading from deciding, own the infrastructure, keep a human in the loop.
Questions people ask
What do I actually receive?
A written report and a phased plan. The report maps your current workflow step by step, flags where the time goes, and marks each step as safe to automate, needs a human checkpoint, or should stay manual. The plan sequences the work into phases with prices, ordered so the first phase pays for itself before the next begins.
What if I do not hire you afterwards?
You keep the plan and you owe nothing further. It is written to be handed to another developer or an internal team. That happens sometimes and it is a perfectly good outcome.
Is the fee really credited?
Yes. If you proceed to a build within ninety days, the full $2,500 comes off the build price.
How much of my team’s time does it take?
Roughly three to four hours total, spread across two weeks. Usually a kickoff conversation, some time watching how work actually moves, a couple of short follow-ups, and a walkthrough at the end.
Do you need access to our systems?
Read-only access helps but is not required, and I will never ask for access to patient records, client files, or anything privileged in order to scope work. Screenshots and a conversation are usually enough to map a workflow.
What if the answer is that we do not need automation?
Then that is what the report says. Sometimes the real problem is a process or staffing problem, and automating it just makes the wrong thing happen faster. I would rather tell you that in week two than six weeks into a build.
Is this an AI readiness assessment?
Yes, though not the survey-and-score kind. There is no maturity rating. Readiness here means specific things: which steps of your actual workflow AI can take on safely, which need a person checking the output, which must stay human, whether your existing agreements already allow it, and what it would cost in phases. Sometimes the honest finding is that you are not ready, or that a step should never be automated, and the report says that.
Two weeks to a real answer.
If a fifteen minute call tells us it is not a fit, that costs you fifteen minutes.