Technology advisory for consequential decisions
Most executives facing a technology decision are not short on opinions. They are short on objective inputs. Vendors have a position. Internal teams have a preference. Consultancies have a statement of work to protect.
I am vested in your success and I am not encumbered. That combination is the whole offer.
Also engaged by PE and growth funds for pre-price technical diligence. For funds
Customers
The problem
Consequential decisions are landing on people who were never hired to make them.
A CEO signs off on an AI budget she cannot independently size. A board approves a platform rebuild on the strength of a deck. A fund partner underwrites a thesis that rests on a technical claim nobody in the room can verify.
None of these are engineering problems. They are business decisions with a technical basis — and the technical basis is the part no one can check.
That is the work.
Start here
Start with the decision, not the org chart.
Pick the one that sounds like your week.
“We’re about to commit real money to AI and we can’t size it.”
You have a number in a budget and a conviction that it matters. What you do not have is a defensible answer to when, where, why, and how much. I build that answer from your business, not from a vendor’s roadmap.
How I size AI investment“Someone is about to look under the hood.”
Diligence, a funding round, an enterprise security review, a board that has started asking sharper questions. You want to know what they will find before they find it.
Technical diligence and readiness“Our pricing isn’t right and we suspect the reason is technical.”
Margin is leaking somewhere between what you sell, what it costs to deliver, and what your systems can actually measure. Usually the pricing problem and the data problem are the same problem.
Pricing and margin architecture“We don’t know our customer as well as we claim.”
Every company says it is customer-obsessed. Few can answer basic questions about behavior, cohort, or intent without a two-week analyst detour. I have built the systems that close that gap.
Customer intelligence“Consequential decisions have no owner.”
No CTO, or a CTO stretched past the point of usefulness. Decisions are being made by default, by whoever is loudest, or not at all.
Embedded and interim leadershipAI investment
Most AI budgets get approved without answering four questions.
The move from a pre-generative to a post-generative world is not a technology decision. It is a capital allocation decision that happens to involve technology. Four questions decide whether the money works:
When.
Is this a now problem or a next-year problem? Moving early on the wrong layer of the stack is more expensive than moving late on the right one.
Where.
Which process, which margin line, which customer moment. “Across the business” is not an answer; it is an absence of one.
Why.
What gets measurably better, and what would falsify the thesis. If nothing would falsify it, you are funding a belief.
How much.
The real number, including the parts nobody puts in the deck — data readiness, integration, model spend at production volume, and the people who maintain it after launch.
I answer these in your terms — your P&L, your customers, your constraints — and I tell you when the answer is not yet or not at all.
The lens
I read technology through the business, not the architecture diagram.
Product.
What to build, what to buy, what to stop. Sequencing that matches how revenue actually arrives.
Pricing.
How your model, your unit economics, and your systems’ ability to measure them either agree or quietly do not.
Experience.
Where the product’s behavior and the customer’s expectation diverge, and what that divergence costs.
Customer intelligence.
What you can actually know about who buys from you, how fast you can know it, and what it takes to make that knowledge routine rather than heroic.
Architecture, security, velocity, and cloud spend all matter. They are inputs. They are not the point.
Proof
Three decisions, and what changed.
The pattern
It tends to show up in three shapes.
Every estimate is “about two weeks.”
Not because the work is two weeks. Because nobody can see far enough to say otherwise.
One engineer knows production
That is not a system. It is a dependency with a calendar.
A security questionnaire triggers a fire drill
Enterprise deals die here, quietly, and you rarely find out why.
None of this is a talent problem. Your engineers are probably good — that is what makes it hard to see.
Capacity
Two engagements at a time.
That is the capacity, and it is deliberate — the value is my attention, and attention does not scale.
Execution does. When an engagement needs specialist depth I did not bring, I pull in people I have worked with directly, under my read and my name. You get a bench without getting a pyramid.
The promise
I am vested in your success. I am not encumbered.
No product to sell you. No implementation revenue waiting on the other side of my recommendation. No incentive to make the engagement longer than the problem requires.
If the assessment says you need a full-time hire, or nothing at all, I will say that.
Do you have a problem, and is there a budget against it?
If yes, that is a short conversation and probably a useful one. If you are not sure yet, that is also a conversation — it just starts further back.
Chicago, IL · San Francisco, CA