The AI Implementation Audit.

At least three actionable fixes you can act on within 30 days, or you don't pay.

"Actionable" is defined on this page, not in marketing language: named against your specific workflows & tools, & executable by the team you already have, without new hires or new spend. That definition goes into the statement of work before you pay anything, so we're both looking at the same bar.

If the audit doesn't clear it, reply "no" to the delivery email & I void the invoice. You don't have to build a case or get on a call to argue about it.

Two weeks. About two hours of your team's time.

Week 1, diagnosis: intake form, three to four stakeholder interviews (30 min each, async-friendly), workflow doc review. End of week: a Compliance-Decay Scorecard showing where your current processes will quietly fall apart under AI, & where they'll hold.

Week 2, the plan: a decision-ready implementation plan & a 90-day roadmap, sequenced & justified. Plus a one-pager written for your boss, because a plan only counts if the person with the budget reads it.

Every audit includes a 30-day check-in call. If you book an Implementation Sprint within 60 days, the full audit fee credits toward it.

See a sample scorecard & a sample boss one-pager, built on a composite organization, at the level of specificity a real client gets.

If your team hasn't deployed anything yet, this is early. The audit diagnoses a rollout in contact with real work, & you don't have one yet. Read the notebook instead, pick one workflow, & come back when there's something to diagnose. I'd rather tell you that now than take $3,500 for a scorecard about a hypothetical.

I won't pick your vendor. I'll tell you what capability the workflow actually needs & what to test for in a trial, & I'll tell you when a tool you're already paying for covers it. The purchase decision stays yours, & it should: I'm not going to be the one living with it.

Sometimes the finding is the process, not the AI. More than once the honest answer has been that the workflow itself is broken & adding AI would make it faster & wrong. When that's the answer, that's what the scorecard says, & the fixes it hands you are process fixes. They still have to clear the three-actionable-fixes bar, & the guarantee still applies.

This is a model, not a measurement of your team. Assume people follow each soft rule in a process about 90% of the time under pressure, which is generous. A process with six such rules runs fully clean roughly 53% of the time. Thirty rules: about 4%.

No single rule is being ignored. What collapses is the chance that any one run follows all of them. The scorecard finds where your workflows lean on that kind of compliance, because those are the exact places AI adoption quietly fails.

I've done this before, at scale, with no budget.

I led an enterprise knowledge-management implementation at Mural Consulting: 14 months, 150+ employees, 8 departments, $0 spend, 99% adoption. The tools were different. The adoption problem was identical.

On the AI side: I built & ran Content Loop to production, a SaaS content system. 7 of 7 podcast episodes processed, 41 content pieces, 100% success after the QA cycle. The QA cycle exists because the early runs failed, which is the part worth knowing about anyone who tells you their system works.

I've shipped a custom MCP server, public on GitHub. I hold 4 Anthropic certifications, proctored & identity-verified. And I run my own business on the same methodology I audit against: an 11-agent production system, documented in a 26-entry Implementation Lab Notebook that feeds the newsletter.