The work
AI transformation is an org design problem, not a tooling one.
The tools are the easy part. What decides whether AI works is the human layer underneath it: how the work is redesigned, how people are led through the change, and whether their expertise still counts. That is the part I build.
What it takes
A redesign, not a rollout.
AI that actually changes how a company works reaches across the whole business, from leadership and change to AI architecture and measurement. Bolt a tool onto the old way and it never gets there.
That is why it has to be designed and led as one connected effort. Handle each piece in isolation and the gains leak away. These are the disciplines it draws on.
How we work together
You don't buy a transformation.
You build it, layer by layer.
Start with one proven win and layer on from there, all the way to a function that runs on AI. Every layer leaves capability behind. Open any piece to see what it includes.
60 days
The AI Opportunity Map
See the low-hanging fruit and the bigger AI opportunities already inside your company, mapped, ranked, and ready to act on.
You can't build the right future on the official version of the present. We map how the work actually flows across the organization, what connects to what, and where the friction and dependencies really live.
What's included
- Conversations with the people doing the work
- How workflows connect across functions
- Where the friction and dependencies live
Most AI implementations do not fail because the agents were built wrong. They fail because the data was never ready.
What's included
- Inventory of systems, data, and integrations
- Gap analysis of what an agent ecosystem would need
- A readiness view before anything is built
Turn everything we found into a plan: know where AI will pay off before committing to a single build. One focused win sells the next three.
What's included
- Opportunity identification and ranking
- Cost and return for each
- A sequenced roadmap you can act on
One quarter
Rebuild a core workflow
Take one high-value process, redesign the future state, place AI where it fits, and build the team's capability to run it.
Everything in the Opportunity Map, plus:
Skip this and teams build faster versions of processes that were already wrong. There is no point process-mapping work you are about to retire. Design thinking asks where you actually want to be, and that changes the destination, not just the speed.
What's included
- Stakeholder and customer interviews
- Facilitated design and prototyping sessions
- A future-state vision that becomes the north star
Deskilling starts in year one, not year five. Physicians using AI diagnostic tools showed degraded accuracy within three months (Lancet). We map what stays human, what is augmented, and what is automated, so people get sharper, not thinner.
What's included
- Human vs. machine mapping, including judgment automation
- Role redesign toward higher-leverage work
- Cognitive sustainability, so expertise grows rather than erodes
The connective tissue of the ecosystem: integrations that remove manual handoffs and let agents work together instead of in isolation.
What's included
- Integrations across your existing tools
- Agent-to-agent orchestration
Validate in real workflows before the whole team sees it. Real conditions, not demo conditions, so every agent is proven before it scales.
What's included
- Pilot design and scoping
- Structured testing and feedback loops
- Iteration before wider rollout
Six months+
Function-wide transformation
Rebuild how a whole function or department works with AI, with champions, measurement, and lasting capability.
Everything in Rebuild a core workflow, plus:
At the individual level, we map what energizes and drains each person in the processes they actually own. Not every high-value process is a big cross-functional workflow; some of the most valuable AI is personal. This is where upskilling and role redesign get pointed at the work that matters most to each person.
What's included
- Work journals: what energizes, what drains
- The individual processes each person owns
- Feeds directly into upskilling and role redesign
A team that cannot think in AI cannot own what gets built for them. Upskilling the whole function first is the difference between a team that uses AI and one that is AI-first, able to keep building on its own.
What's included
- Foundational fluency across the function
- Prompting, evaluation, and testing discipline
- A framework that removes limiting beliefs and builds real confidence
Most firms introduce champions at rollout, as messengers. Real champions are co-designers from day one: in the interviews that shaped the design, first to test, credible because they lived it.
What's included
- Champions identified during discovery
- Onboarded as co-designers, not communicators
- Managed through build, testing, and rollout
Purpose-built agents designed around the roles and practices defined earlier, each fully documented so your team can maintain and extend them long after I leave.
What's included
- Purpose-built agents for the team's key functions
- Full documentation: purpose, logic, and construction
- Champion-tested in real workflows first
Resistance is usually misdiagnosed as a communication problem. For people with real automation exposure, wariness is a rational response, and treating it as messaging produces resentment, not adoption (NOBL).
What's included
- A change-management framework built for the engagement
- Communications aligned to the real story
- A change manager supporting the team throughout
AI done without intention creates silos; people talk to tools instead of each other, and the informal moments that generate team intelligence quietly disappear. Practices that prevent that have to be built in from the start.
What's included
- Rituals that keep humans connected
- Norms for what to keep human and what to hand over
- Team intelligence protected, not eroded
Productivity metrics tell you if output went up, not if the people producing it are getting better or worse (NOBL). At function scale, we measure what actually predicts whether transformation holds.
What's included
- Pre and post survey design across the function
- Cognitive load, energy, and quality of judgment, alongside productivity
- A clear impact readout
Ongoing
Org-wide transformation
Scale what worked in one function across every department, with the operating model and champions to sustain it.
Everything in Function-wide transformation, plus:
76% of executives believe their teams are enthusiastic about AI. The real number is 31% (BCG). The gap lives at the leadership layer. When leaders visibly model AI use, employee trust in it jumps 30 points (Microsoft, 2026).
What's included
- 1:1 and group coaching for leaders and managers
- The four-layer model: who you are, how you learn, how you relate, what you build
- Modeling the learning, not performing certainty
Automation that works across functions, not just inside one, so the whole system moves together instead of creating new islands of productivity.
What's included
- Cross-functional automation strategy
- Orchestration across teams and tools
What scales is not the technology, it's the organization's capacity to keep building. Champions lead ongoing iteration in their functions, and the gains compound.
What's included
- Champions network across the organization
- An operating model for continued build
- Compounding gains, function by function
Rooted in research
Every choice here is backed by the evidence,
not a hunch.
Stanford HAI · AI Index →
Adoption is everywhere. The measurable gains cluster in the organizations that redesigned the work, not the ones that only bought the tool.
NOBL · Work Redesign →
Durable change comes from rebuilding how work flows — layering new tools onto old processes reliably fails.
Gallup · State of the Global Workplace →
Engagement drives performance, and disengagement costs the global economy trillions every year. Capability has to be built into people, not around them.
Microsoft · Work Trend Index →
People are already bringing their own AI to work — faster than most companies can guide it. The question is no longer whether, but how well.
BCG →
Roughly ten percent of AI value is the algorithm and twenty percent the technology. The other seventy is people and process — exactly where I work.
Google · Project Oxygen →
The strongest results trace to how work and management are designed — not to raw technical skill. Design is the differentiator.
Deloitte →
Whether AI investment returns anything comes down to human capability and trust — the softest inputs decide the hardest numbers.
The Lancet →
Even in the hardest human domains, outcomes improve when the tool augments expertise rather than replacing it.
Questions I get
The things people ask before we start.
What does an AI consultant actually do?
An AI consultant helps a company actually adopt AI, not just buy tools. That means finding where AI genuinely fits, redesigning how the work gets done, leading people through the change, and measuring the result. I focus on the human side of that, which is where most AI efforts stall.
Where should we start with AI?
Not by mapping your current processes in detail. That tells you about today, and AI is about where you're going. I start by getting the real pain points and the must-haves on the table, then defining what the work should become. That is where the money is well spent: designing the future state, not documenting the present one.
Why hire one AI consultant instead of specialists or a big firm?
Because the two obvious choices both break. Hire a single specialist and you get one slice of a job that actually needs a whole stack of disciplines: business acumen, organizational redesign, process improvement, design thinking, change management, leadership coaching, communications, building the AI, and measurement. Hire a big firm and you get the breadth, but they lose sight of it, juniors do the work, and no one owns the outcome. I am the third option: one senior person who holds the whole stack, architects the whole, and brings in specialists only as the work needs more depth or scale. Broad enough to see the whole picture, close enough never to lose it.
Will AI replace my team's jobs?
It doesn't have to, and the companies pulling ahead aren't the ones cutting. AI absorbs the routine so your people can move to higher-value work. Done well, adoption grows your team's capability instead of shrinking headcount. That is the whole approach I build, and I measure it.
How do I know if my company is ready for AI?
Readiness is less about the technology and more about your people and processes. A quick way to gauge it is my AI Readiness Assessment: a few questions that show where you are strong and where the gaps are. From there, we would talk through what to do first.
What is AI organizational design?
AI organizational design is rethinking how work gets done when AI is part of the team, rather than bolting tools onto the old structure. It covers how roles are shaped, how people are led, how decisions flow, and how you measure whether it is working. It is the human layer underneath any AI rollout, and it is roughly seventy percent of where the value actually comes from.
What's the difference between AI strategy and AI implementation?
Strategy is the plan; implementation is making it real inside how people actually work. Most consultants sell one or the other. I do both, because a strategy that never gets executed changes nothing, and tools without a plan just add noise.
What size companies do you work with?
Everything from a single founder to a 37,000-person enterprise. The principles are the same; the scale of the rollout changes. I work especially with mission-driven companies and B Corps.
Do you speak at events?
Yes. I give keynotes and workshops on AI, the Future of Work, and the human side of adoption. See the Speaking page, or get in touch.
Let's talk
Tell me where the work is getting stuck.
Every engagement starts with a conversation about the real problem, not a pitch. That is the fastest way to know if I am the right person to solve it.