What a two-week AI audit actually looks at

Every engagement we take starts the same way: two weeks inside your operation, a fixed fee, and a ranked answer on where AI pays — before you commit to building anything. Here is what we actually do with those two weeks.

Most AI proposals start with a solution looking for a problem. The audit exists to kill that dynamic. We arrive with no architecture to sell you, and we leave with a document you own either way: a ranked map of the workflows in your business where automation pays, what each one costs you per year, what it would take to build, and the order to do them in.

Days 1–4: workflow mapping

We sit with your operators — the people who actually touch the work — and map the workflows step by step. Quoting, scheduling, invoicing, order intake, dispatch, document handling. Not the process as the org chart describes it; the process as it actually runs, including the sticky notes, the re-keyed data, and the spreadsheet someone maintains at 9pm.

The output is unglamorous and essential: for each candidate workflow, the volume of transactions, the hours they consume, the error rate, and where the errors land. Most owners have a feel for this. Almost none have it on paper with numbers attached.

Days 5–8: data readiness

Every workflow depends on data that lives somewhere. We check what shape it is in: where prices, schedules, and customer records actually live, how current they are, and what it would take to connect them. A workflow can be a perfect automation candidate on paper and a dead end if its inputs live in three inboxes and a filing cabinet. Better to learn that in week one than month three.

Days 9–10: ROI ranking

Each opportunity gets a payback estimate in hours and dollars, weighed against build cost and difficulty. We show the math, and we mark every assumption. The result is a ranked list, not a vision deck — usually two or three workflows where payback lands in months, and a clear view of what to leave alone.

The map is yours. If you never build with us, you still know more about your own operation than you did two weeks ago.

If you continue, the audit fee credits in full toward the pilot, and the pilot scope comes straight off the map — no re-discovery, no new slideware. If you don’t, the map is yours. That is the entire trick: the audit is cheap enough to be a rounding error against a wrong build, and honest enough to be worth acting on.