AI audit & readiness

Find Out Where AI Pays Off Before You Spend on It

Everyone is telling you to "add AI". Almost nobody will tell you where it actually earns its keep in your business, and where it is an expensive distraction. That is the audit: a written, costed plan you own, whether or not you ever build with us.

What an AI audit actually is

An AI audit is a structured review of how your business actually operates — the workflows, tools, data and repeated manual work — that identifies where AI would genuinely pay for itself and where it would not. The output is a written report: a ranked list of opportunities, what each would cost to build, what it would save, and which to do first. It is a plan you own and can take to any developer, not a sales document that only makes sense if you hire the people who wrote it.

The reason it is worth doing before building anything: the most common failure mode in business AI is not bad technology, it is automating the wrong thing. A week of looking properly is the cheapest insurance against a quarter of building the wrong system.

What you get

A Report Worth Showing to Someone Who Isn't Us

Fixed price, agreed before it starts. One to two weeks for a single team. Everything in writing.

AI Audit & Readiness Assessment

We interview the people doing the work, look at the tools and data they actually touch, and test today's models against your real examples — your emails, your documents, your tickets — rather than a demo. Then we write down what we found, including the opportunities that are not worth pursuing and why. Rankings are by payback, not by what would be most interesting to build.

What's in the report

  • A map of your current workflows, tools and data
  • Ranked AI opportunities, each with build cost and expected saving
  • Tests of current models against your real examples, with results
  • The data gaps and risks that would block each opportunity
  • What to buy off the shelf versus build custom
  • A recommended order of work with fixed prices for each step

Evaluated against

ClaudeChatGPTGeminiWhisperZapiern8nYour stack
Get a fixed price for your audit

What an audit typically finds

Patterns repeat across businesses. These are the ones that come up most, roughly in order of how often they pay for themselves.

FindingUsual verdictWhy
The same questions answered by email every weekAutomateA grounded chatbot or a better FAQ page, depending on volume
Data retyped between two systems by handAutomateUsually the highest payback in the whole report, and the least glamorous
Documents read, summarised and filed manuallyAutomateCurrent models handle this well when the output is reviewed, not auto-sent
Decisions that depend on one person's judgmentDon't automateAI drafts, human decides. Removing the human here is where the horror stories come from
"We should have a chatbot" with thin content behind itFix content firstA bot over stale documentation produces confident wrong answers faster than a human could
The honest part

Some audits conclude the best next step is fixing a process, updating content, or buying an off-the-shelf tool rather than building anything. When that is the finding, the report says so, with the reasoning written down. An audit is only worth paying for if its recommendations survive being shown to someone who isn't us.

How it runs

One to Two Weeks, Ending in a Written Plan

A few interviews, access to the tools involved, and real tests against your own examples.

  1. 01

    Interview the People

    Short conversations with the people doing the work, not just the people managing it. The repeated, boring tasks nobody thought to mention are usually where the money is.

  2. 02

    Map the Systems and Data

    What tools you run, where the data lives, and what state it is in. Half of "AI readiness" is just whether the data an AI would need actually exists somewhere reachable.

  3. 03

    Test Against Real Examples

    We run current models against your actual documents, emails and tickets and record the results, so every recommendation in the report is backed by a test, not a vendor's claim.

  4. 04

    Deliver the Plan

    Ranked opportunities, costs, savings, risks, and a recommended order of work with a fixed price against each step. Yours to keep, build with us, or take elsewhere.

Common questions

About the AI Audit

A structured review of how your business actually operates — its workflows, tools, data and repeated manual work — that identifies where AI would genuinely pay for itself and where it would not. The output is a written report: ranked opportunities, what each would cost, what it would save, and which to do first.

It is fixed-price, agreed before it starts, and depends on how many teams and systems are in scope. A single-team audit is a small engagement measured in days, not months. Book a free consultation and you will have a written number by the end of the call.

A written report: a map of your workflows and tools, ranked AI opportunities with build cost and expected saving for each, results of model tests against your real examples, the risks and data gaps that would block each one, and a recommended order of work. The report is yours to take to any developer, including one that is not us.

No. Some audits conclude the best next step is fixing a process, updating content, or buying an off-the-shelf tool rather than building anything custom. When that is the finding, the report says so, with the reasoning written down.

Typically one to two weeks from kickoff to written report for a single team or department: a few interviews, access to the tools and data involved, and time to test current models against your real examples.