Phase 1
Find out what AI is worth to your firm before you buy any.
Two weeks. A fixed fee, quoted on the size of your business and credited in full against implementation if you proceed. We find the AI already in use, talk to the people doing the work, rank what is actually worth automating, and test all of it against the standards you are held to. You keep the report either way.
Most firms buy the tool before they know the job.
The usual sequence is backwards: pick a product, roll it out, then discover which work it actually fits and which rules it just broke. Two weeks of looking first is cheaper than a year of a deployment nobody uses — and it is the only way the second question, what to implement, has a defensible answer.
Four things happen in the two weeks.
What is already running, whether or not anyone approved it.
Every firm we have looked at is already using AI somewhere. We find where: which tools, on whose accounts, touching which client information. This is a factual inventory, not a reprimand — you cannot govern a tool you do not know is in the building.
Where the hours actually go, from the people spending them.
We talk to staff across roles, not just the partners. Partners know what the firm bills; the people preparing the work know which parts are repetitive, which reviews get redone, and where they have already started improvising with AI on their own. That gap is usually the most useful thing we find.
What is worth automating, ranked — and what is not.
We map the work against three things: how many hours it consumes, how repeatable it is, and how much judgement it carries. Work that is high-volume and low-judgement ranks first. Work that carries professional judgement is marked to stay with a human, explicitly. The output is an ordered list you can argue with.
Findings against the standards that already bind you.
Each finding is tied to the rule it falls under and paired with a remediation. For a CPA firm that means IRC §7216, GLBA and your WISP, Circular 230, and AICPA confidentiality. For firms in other regulated work we score against the obligations you actually carry.
What you leave with.
Every engagement ends in the same document structure, so it can be compared across firms and across time. Walk a sample report before you commit to anything.
A baseline Implementation Skoor
One number, scored against a published rubric, for where the firm stands today. It is a starting line, not a grade — it exists so the second measurement means something.
The ranked opportunity map
The ordered list of work worth moving to AI, with the reasoning shown, and the work we recommend leaving alone.
Findings and remediations
Every exposure we found, tied to the standard it falls under, each with the specific change that closes it.
A priced implementation plan
What it would cost to do the work, sequenced. If you proceed, the assessment fee comes off it in full. If you do not, the plan is still yours.
What it is not.
- A software trial. Nothing has to change in your stack during the two weeks.
- An attestation or a regulatory filing. It is our assessment against published standards, not an opinion you can hand a regulator.
- Contingent. The report is yours whether or not you continue, and we will tell you if the honest answer is that you do not need us yet.
The rubric we score against is published, not proprietary. Read it here — you should be able to see how a number was reached before you accept it.
And then, only if it is warranted.
The assessment stands on its own. If the findings justify implementation, the plan is already written and priced, and the fee you paid comes off it in full. If they do not, you have a written baseline and an ordered list of what to revisit — which is worth having regardless of who does the work.
See how the full engagement works or what this looks like for a CPA firm.
Two weeks to a defensible answer.
Fixed fee, quoted on the size of your business, credited in full against implementation. The report is yours either way.