AI Readiness Assessment
Two to three weeks. One document your CFO will actually read: where AI pays in your business, what it would cost, what would stop it, and the payback in your numbers, not an analyst's.
From $7,500, fixed price
Established organizations that want a number before a project.
Manufacturers, distributors, banks and credit unions, professional firms, healthcare administrators and public agencies, typically $5 million to $100 million in revenue. You have real processes, people who know them, and constraints that a generic AI pitch ignores: regulators, customers' data, union agreements, auditors.
You are hearing that AI could help and you want to know where, by how much, and what it would take, without committing to a build first.
A systems and data inventory
What you have, where it lives, and how to get data out of each system: API, export or neither.
Three to five scored opportunities
Each scored on the four levers below, with a conservative, expected and upside case built on fully loaded labor cost, not salary.
The governance gaps
The security, privacy and compliance issues that would block each opportunity, and what closes them.
A roadmap with total cost
Build, model usage, cloud and your own staff time, with a payback period for each item. Written so any vendor can execute it.
A readout with your sponsor
One session to walk the findings, answer the hard questions and agree on what, if anything, happens next.
Four levers. Every opportunity sorts into them in a minute.
We never claim eighty percent time savings. We model thirty to fifty percent on the task with a person still in the loop, show the number at thirty, and present three cases. CFOs trust people who show the low case.
| Lever | What it means | How we measure it |
|---|---|---|
| Labor hours reclaimed | Time a task takes now against after | Hours a week × fully loaded hourly cost × 50 weeks |
| Cycle time | How fast a quote, order, claim or answer moves | Days removed × what speed is worth: win rate, cash flow, penalties avoided |
| Error and rework reduction | Mistakes that cost money downstream | Error rate × cost per error × volume |
| Revenue capacity | Doing more with the same people | Additional quotes or orders handled × margin |
The total always includes everything: our fee, model and API usage, cloud, and your staff's time for training. Then payback period and twelve-month net, in the metrics your business already uses.
Measure first. Estimate second.
Week one: listen and count
Interviews with three to five people who do the work. The systems inventory. We hand you a counting sheet and you measure the current state for a week: how many documents, how many hours, how many errors. Half the value of the assessment is that nobody has measured this before.
Week two: score
Three to five opportunities scored on the four levers, three cases each, and the governance review of what each one would need: identity, logging, data handling, regulatory hooks.
Week three: roadmap and readout
Total cost, payback and sequencing. A readout with your sponsor. A short third week is usual; a shorter engagement when the inventory is small.
- A sponsor who can make a decision at the end.
- Three to five people for one-hour interviews.
- Read access or exports from the systems in scope, under NDA.
- One week of counting with the sheet we give you.
From $7,500, fixed price. The price is fixed at signature based on the number of systems and people in scope. Half is invoiced at signature and the balance on delivery of the report, net 15. No work starts without a signed statement of work.
Before you ask.
Do we need to clean up our data first?
No. Messy data, missing fields and three versions of the same spreadsheet are findings, not blockers. The inventory tells you which problems matter for which opportunity, and which ones you can ignore.
What if nothing turns out to be worth automating?
Then the report says so, with the counts to back it up. You keep the systems inventory, the baseline measurements and the governance review, which are worth having regardless. We would rather tell you that than sell you a build.
Do we have to hire you to build what the roadmap recommends?
No. The roadmap is written so any competent vendor, or your own team, can execute it. Most clients who proceed do so with us because we already know their systems, but that is their choice to make at the end, not a condition at the start.
Where does our data go during the assessment?
It stays in your systems. We work from read access or exports under a signed NDA, inside your environment where possible. Nothing is sent to a model provider during the assessment without a written decision from your sponsor.
Who needs to be involved on our side?
A sponsor who can make a decision at the end, three to five people for one-hour interviews, and someone who can pull exports or grant read access. Expect about a day of your team's time in total, spread over the engagement.
Can it be done remotely?
Yes. Interviews and the readout work well on video, and the counting sheet is a shared spreadsheet. We come on site when seeing the floor or the office changes what we would recommend.
Start with the assessment.
Fixed price, two to three weeks, and a document you can hand to your board or to any vendor.