Four questions, answered with evidence rather than opinion.
-
Where are we actually exposed?
What a competitor with better adoption can do to you inside twenty-four months — and how long you have before that gap becomes structurally expensive to close.
-
What should we do first?
Not a list of everything possible. The two or three things that matter, sequenced, with the reasoning visible.
-
What will it cost, and who does it?
Real numbers, real roles, real timelines. Every recommendation carries an owner and a prerequisite.
-
What are we already carrying?
Governance, data, vendor and regulatory risk — including the tools your people are already using that nobody approved.
Six dimensions, scored one to five.
Built from eight to twelve structured interviews across leadership, operations, IT and frontline staff, plus a document review. Not a survey. Not a maturity model you fill in yourself.
Strategic clarity
Does leadership agree on what AI is for here — or is there a stated thesis standing in for anxiety?
Data foundation
Does the data the obvious use cases require actually exist, and is it accessible, accurate and governed?
Process readiness
Are the target processes stable enough to automate — or would you be automating chaos?
Talent and capability
Who internally can own this, and how wide is the gap between that person and what the work needs?
Governance and risk
Policy, approval paths, vendor diligence, regulatory exposure — and what staff are already doing unsupervised.
Capital and commitment
Is there budget and sustained executive attention, or is this a board asking for a memo?
Readiness assessment
Twenty to thirty pages. Findings and scores across all six dimensions, with the evidence behind each one.
Prioritized roadmap
Twelve to eighteen months, phased. Owner, cost range, prerequisite and success measure on every item.
The kill list
What not to pursue, and why. Clients consistently report this as the page that saves the most money.
Board presentation
Delivered live, written for directors who are not technologists. Plus a two-page standalone summary.
Board briefing
Two working sessions and a written primer. Enough shared vocabulary that the board can have the real conversation instead of talking past itself.
ForA board that needs to get literate before it can decide anything.
Readiness diagnostic
The full assessment. All six dimensions, the roadmap, the kill list, and a live presentation to the board or leadership team.
ForAn organization that has decided to act and needs to know where to start.
Multi-unit diagnostic
The diagnostic run across several business units or sites, with comparative scoring and a single consolidated roadmap.
ForA holding company or multi-site operator with uneven readiness.
Every engagement is fixed-scope and fixed-fee, quoted in writing before any work begins. No hourly billing, no open meter, and no change orders unless you change the scope. If the work turns out to be larger than I estimated, that is my problem rather than yours.
I was Chief Sustainability Officer of Invitae, a NYSE-listed genetics company, where I built out the sustainability and ESG function and the governance around it — board-level oversight, disclosure controls, and the reporting mechanisms a public company gets judged on. I reported to the COO and presented regularly to the board. I've been on the receiving end of this kind of assessment, which is most of what taught me the difference between one a board can act on and one that just gets filed.
Before that, chief of staff roles at Amazon across fulfillment and delivery operations, and several years at EY in advisory — running client accounts and leading post-merger integrations and finance transformation programs for Fortune 100 companies. Earlier, at Accenture, I was on the delivery end of that work: enterprise system deployments across SAP, point-of-sale and workforce management, through go-live and the weeks after. I've been the one accountable when a system finally meets a working operation, which is where most implementation plans discover what they got wrong.
Today I lead a team that builds and ships production AI agents inside a large enterprise, along with the telemetry and governance that determine whether any of it actually gets used. That vantage point is most of why I'm useful here. I get to watch the full distance between an AI pilot that demos well and one that survives contact with a working operation — across a lot of them, in a lot of different conditions — and the patterns repeat.
That combination is the practice. I've sat on the board-reporting side of a public company, the delivery side of a Big Four advisory firm, and the build side of enterprise AI. Organizations rarely fall behind on AI because they chose the wrong model or the wrong vendor. They fall behind because nobody with standing looked at the whole picture and said plainly what was worth doing and what wasn't.
I'm not a firm and I'm not trying to become one quickly. You work with me. That limits how much I take on, which is deliberate — I'd rather run four engagements a year properly than twelve badly.