Work · Fractional partnership · 2026
Quarterly AI budgets yield quarter-good AI.
AI transformation is a decade-long change-management project. Most companies budget it like a quarterly initiative — and get the AI they paid for. I work with CEOs who see the longer game: building the operational muscle that turns AI from a line item into operating leverage.
01
The math nobody wants to talk about
AI adoption is up. AI ROI isn’t. The gap isn’t where most people look.
Enterprise AI data from 2025 and 2026 tells a consistent story: companies are buying AI faster than they’re getting value from it. Different firms stall in different places — but the patterns repeat.
Where enterprise AI stalls — industry data, 2025-2026
Pilots that don’t scale
~85%
Reach measurable ROI
~30%
Different companies have different bottlenecks. Legacy firms usually stall on change management — the data exists, the budget exists, the organizational muscle to adopt new ways of working doesn’t. Mid-market firms usually stall on data — ambition outruns infrastructure. Almost everyone stalls on the pilot-to-production gap.
The successful minority have one thing in common: they treat AI as multi-year infrastructure, not quarterly initiatives. They invest in the muscle, not just the tooling.
The vendors talk about the model. The work is everywhere else.
02
In a decade, every employee will manage an agent
Not use one. Manage one. Guide it. Evaluate it. Grade it. Decide when it’s right, when it’s wrong, when to escalate, when to retire it.
That’s a different organizational design than the one most legacy companies are built around. It requires:
→
Leaders who set the agenda for what gets automated — and what stays human.
→
Operators who design the workflows where employees hand off the grunt work — writing, calculating, summarizing, following up — while keeping the core of their job.
→
Employees who treat AI as a teammate they manage, not a tool they use.
→
A culture that grades systems against outcomes, not adoption rates.
That’s a mindset shift, not a tooling decision. It’s the kind of change that takes years to build and can’t be downloaded from a vendor. The companies that build this muscle in 2026-2028 will operate at a different cost basis by 2030.
The ones that don’t will be acquired by the ones that did.
03
The gap I close
Most AI consultants come from data science or software engineering. They’re excellent at the 10% — picking models, building pipelines. They’ve usually never run an operation.
Most operations consultants come from process improvement or management consulting. They’re excellent at workflows — Six Sigma, lean transformation, change rollouts. They usually don’t understand AI deeply enough to design around it.
The two rarely cross-pollinate. What you actually need is someone who can:
→
Read a P&L and tell you where AI fits — not in slides, in line items.
→
Design the operational loop where humans hand the grunt work to AI and AI hands the edge cases back.
→
Build the orchestration layer that makes the system self-correcting after deployment, not just at launch.
→
Ship the proof against real capital — not slides — before recommending it to yours.
04
What I actually do
Fractional
CAIO
Chief AI Officer, part-time
Strategic AI leadership for mid-market companies that can’t justify a full-time CAIO. Board-ready roadmaps, vendor selection, hire-or-build calls, risk framing. The leader who sets the agenda for what gets automated.
10–20 hrs/week · 3–6 month minimum
Operational
AI Design
Workflows that ship and improve
End-to-end design of the operational loops where humans hand grunt work to AI: agents, evaluation, drift detection, calibration, human-in-loop gates. Delivered as a working reference architecture integrated with your stack — not a slide.
Fixed-scope · 6–12 weeks
Build
Alongside
An architecture co-pilot, embedded
For engineering and product teams with AI in production who need orchestration discipline. I pair with your leads on system design — orchestration patterns, operational loops, calibration cadence. The work is upstream of code: the decisions that shape what gets built. Not code review.
Embedded · 4–12 weeks
05
Who I work with
I work with CEOs, COOs, and operating partners who see the decade-long bet before their peers do.
Companies and leaders I partner with usually share three traits:
→
They want operating leverage, not press releases. The goal is a different cost basis, not a media moment.
→
They’re willing to invest in change management alongside technology — not just buy tooling and hope adoption follows.
→
They treat AI transformation as multi-year infrastructure — the way they treated cloud migration, or ERP rollouts, or any other change that reshaped how their organization works.
If any of that resonates — or if you’re weighing whether the path you’re on is the right one — the engage form below is the place to start. A short note about your situation is enough.
06
How we work together
Every engagement runs through three phases. The cadence and depth scale to fit your situation, but the shape is consistent.
01
A short, intense diagnostic. Where AI is today in your organization (audit). Where the change-management muscle exists and where it doesn’t. Which loops in the business are real candidates for AI handoff. Output: a written assessment with three to five concrete recommendations.
02
Design the operational loops, the orchestration layer, the calibration cadence. Sequence the change-management interventions. Output: a working reference architecture integrated with your stack, and a 90-day adoption plan with named owners.
03
Operating cadence
Months 3+
Ongoing fractional CAIO support. Weekly or bi-weekly cadence. Calibration reviews, vendor decisions, board reporting, the operational loop running on autopilot. Output: a system that improves on its own and a team that can run it without me.
07
Engage
Inquiry
Two engagement slots open. What are you trying to ship?
A short note about your situation is enough to start — what you’re trying to change, where you are today, what success looks like. I reply within 48 hours; a 30-minute scoping call follows if there’s mutual fit.