AI Operations Assessment
Find out what your current process actually costs you, before you automate it.
A structured review of where AI and automation can save you time and money, and where they would create risk if you deployed them wrong. Built for businesses in the $1M to $10M range. You get a written report you own outright, with a number attached to every recommendation.
From $5,000, scaled to the size of your business. Half on signing, half on delivery. Half the fee comes back as credit toward a build.
The smaller first step
One broken process? That is a $750 question, not a $5,000 one.
The Process Diagnostic is one 60-minute call focused on a single process that is costing you time or money. You get a two to three page written verdict: how the process works today, what it is costing you per month, and whether AI or automation can actually fix it, before you spend anything on a build.
What it deliberately is not: no build proposal, no security and governance layer, no roadmap. One process, one clear answer, nothing more. If the answer is that a $200 setting change fixes it, or that automation will not pay for itself, that is what the document says.
Bottom line: the Diagnostic answers one question. The Assessment below maps the whole operation. Start with the size of your actual problem, not the size of the engagement.
The Diagnostic is right if
- →You already know which process is broken and want a straight answer on fixing it
- →You want a verdict in writing before committing real money to anything
- →You want to see how I work before a larger engagement
- —If it is the whole operation that needs the review, keep reading: that is what the Assessment below is for.
Why this exists
Most businesses adopt AI without ever calculating what the old way was costing them.
Which means they cannot tell whether the new way is better. They have a tool, a monthly bill, and a feeling. No baseline, no comparison, no way to know if the thing they bought is earning its keep.
This assessment starts with your numbers. How long the process takes now, how often it runs, who touches it, and what that adds up to per month. Everything after that is measured against a figure you can verify, because it came from your own operation.
Bottom line: you should know the cost of doing nothing before anyone quotes you the cost of doing something. That number is the first thing this report gives you.
What you get
- →Discovery call analysis and a current-state map of your processes
- →A quantified cost of inaction, calculated from your own numbers
- →A proposed solution architecture for the highest-impact opportunities
- →A security and governance layer, covered in detail below
- →An implementation roadmap and a prioritized action plan
- →Quantified ROI, plus two feedback sessions to refine the report
- →Delivered as a written report you own outright
The part nobody else includes
The security and governance layer
Five documents that answer the question most AI projects never ask: when this goes wrong, who is accountable and what happens next. This is the section that exists because I spent sixteen years reconstructing failures after the fact.
01
Decision framework
Which tasks should use AI, which should use plain automation, and which should stay with a person. Written down, so the next decision is not a coin flip.
02
Tool risk map
Every tool touching your data, what it can reach, and where credentials live. The map you wish you had before someone asks where the customer list went.
03
AI acceptable use policy
Written, in plain language, so your team stops guessing what data and what tasks are off limits.
04
Verification protocol
How AI output gets checked before it reaches a customer, a decision, or a record. Verification takes more expertise than generation, not less.
05
Human-in-the-loop standard
Who reviews what, at which point, and what they are accountable for. AI does not carry accountability. Your business does.
—
And it is yours
These are your documents. They survive this engagement, they work with tools I had nothing to do with, and they do not expire when we stop working together.
How it runs
Three stages, and you see the report twice before it is final
01
Discovery and mapping
Calls to understand how the work actually happens, not how the org chart says it happens. Current-state mapping, and the numbers behind it.
02
Analysis and drafting
Cost of inaction, solution architecture, the governance layer, the roadmap, and the ROI figures. Written up as a draft report.
03
Two feedback sessions
You read it, you push back, it gets refined. Twice. A report you disagree with is not worth what you paid for it.
Terms worth knowing up front
Half the fee comes back if you build
Assessment clients receive a credit equal to 50% of the assessment fee, applicable toward a build engagement of $5,000 or more, valid for 90 days from report delivery. Whatever tier you are on, proceeding to a build halves what the assessment cost you.
That is deliberate. The assessment exists to tell you what is worth building, so it should not be a toll on the way to building it. It also means I have no reason to recommend work that is not worth doing, because the credit applies either way.
And if the answer is don't: some processes should not be automated, and some businesses are not ready. If that is what the numbers say, that is what the report says. You still own it, and you still know something you did not know before.
What is not included
- —The actual build or implementation. That is scoped separately, and the credit applies to it.
- —Software or tool subscriptions. You buy those directly, on accounts you own.
- —Ongoing support or maintenance. That is a separate retainer, starting at $500 a month.
Payment
- →50% deposit due when the engagement letter is signed, before work begins. It reserves the engagement and is non-refundable.
- →The remaining 50% is invoiced on delivery of the completed report.
Right fit
This is the right starting point if
- →You do roughly $1M to $10M a year and have more repetitive work than your team can absorb
- →You are being pitched AI tools and cannot tell which ones are worth it
- →Something is already running that nobody is formally checking
- —If you already know exactly which one process you want fixed, the $750Process Diagnosticis the cheaper answer
- —If missed calls are the whole problem, start withClearLineinstead
- —If you want AI running with nobody reviewing it, this is not built for that
Twenty minutes is enough to know whether the assessment is the right starting point.
Free. No pitch and no deliverable. If a $750 diagnostic would answer your question instead, I will tell you that on the call.
Pisano Designer · Radu Pisano