Most AI engagements end with a document. These are the three services that turn one into something your business can actually feel — and you don't need all of them.
The Whole-Company AI Review identifies what matters and in what order. These three services are how the plan gets carried out — individually, or in whatever combination the review recommends.
You can also engage any of these directly. If you already know what you need built, or which team needs training, there's no requirement to run a review first — I'll tell you honestly if one would change the answer.
Your people are already using AI. The question is whether they're using it well, and whether anyone has told them what's acceptable.
A practical session for owners, managers, and staff. What these tools can and can't do, how to spot a good use case, how to judge whether output is trustworthy, and what should never be pasted into a chat window.
Two hours. Whole company or by department.
Role-specific sessions built around your actual workflows. Sales, operations, finance, client service — each team working on the real tasks they do every week, not generic examples.
Scoped to the teams and workflows that matter most.
A demo of whatever tool is popular this month. Tools change; judgment doesn't. The training is about knowing when to use AI, when not to, and how to tell whether what came back is right — which outlasts any particular platform.
When the opportunity is worth testing, I build a working application — not a mockup, not a specification for someone else to implement.
The most common request, and usually the most valuable. A dashboard that pulls from the systems you already run and shows the five to ten numbers that actually drive the business — instead of a spreadsheet somebody rebuilds every month.
The workflow that eats hours every week. Intake checking, proposal drafting, reconciliation between two systems that refuse to talk — built to fit how your team actually works.
Portals, self-service tools, and new digital services. Built small first so you learn whether customers want it before committing to the full version.
Mindbody, Stripe, HubSpot, QuickBooks, Google Workspace, Zapier, and most major platforms — plus custom databases and anything with an API.
If your systems don't talk to each other, that's usually the real problem — not the absence of AI.
Typically two to four weeks depending on data complexity. Scope and cost agreed before anything starts.
The application is yours to keep, with documentation. No lock-in, no ongoing licence.
Most of what goes wrong with AI happens after the decision. The tool gets bought and half-used. The workflow reverts to what people did before. The person who championed it moves on, and six months later nothing has changed except the invoice.
Agreeing what a change was meant to improve, and how we'd know — before it ships
Watching adoption where it actually stalls: the workflow, not the tool
Evaluating new tools and vendor proposals before you commit
Policy and responsible-use questions as they come up
Deciding what's worth doing next — and what isn't
Advisory is the part that keeps it honest. Before anything ships we agree what it was supposed to improve and how we'd know. Then I stay in it while it's actually being adopted — a standing monthly session, and access in between when something comes up — looking at whether it's working, what to adjust, and what's worth doing next.
Month to month, with no annual commitment. Most clients start after a review, once there's a plan worth maintaining.
If it's working and you don't need me, I'll say so.
You don't need all three on day one. Most companies start with whichever the review ranked first, see whether it delivers, and decide from there — at your pace, on your budget.
Whichever opportunity carries the clearest value and the fewest dependencies. Something you can judge honestly within weeks.
Against the measure agreed beforehand, not a general impression. If it didn't deliver, that's worth knowing before the second investment.
Move to the next item, deepen the first, or stop. Every engagement is scoped so stopping is a real option rather than an expensive one.
A custom build meant six-figure budgets and six-month timelines. For a company between $1M and $20M in revenue, that math almost never worked — so you bought software designed for someone else and adapted around it.
AI-assisted development changes the arithmetic. I can build in weeks what used to take months, which makes purpose-built software worth considering for problems that previously had to be lived with.
Faster development doesn't mean less rigor. Scope is agreed in writing, the work is tested, and you get documentation.
You own what's built. No proprietary platform, no dependency on me.
And the hard part was never the code. It's knowing what the software actually needs to tell you — which is why the review usually comes first.