Operating principle
I work with mid-market executives and investors when AI decisions are high-stakes, the information is incomplete, and internal incentives make clarity harder than it should be.
The core value is independence.
I'm not paid to push tools, vendors, or implementation. I'm paid for decision quality and execution oversight.
Boundaries (non-negotiable)
I don't:
- • Write code or build AI systems
- • Act as an agency, vendor, or delivery team
- • Embed full-time in one company (I work with multiple clients)
- • Do AI work to satisfy internal politics or "innovation theater"
I do:
- • Build strategy and roadmaps
- • Validate tools and vendors (stress-test claims, not just compare feature lists)
- • Train leadership and teams (from fundamentals to specific tools)
- • Oversee execution (manage vendors, contractors, internal teams)
- • Establish governance (so decisions stay aligned and don't fragment)
- • Provide technical due diligence (for investors evaluating AI initiatives)
If you need someone to build software, I'm not the right fit. If you need someone to make sure you build the right software — with the right partners, in the right order — that's what I do.
The decision and execution loop
Every engagement follows this process:
Frame the decision and constraints
Most leadership teams arrive with 5–10 "decisions" that are actually 1 decision with unclear constraints.
I start by isolating the actual decision from the surrounding noise.
In the first session, I force:
- • A single, explicit decision statement
- • Hard constraints (budget, timeline, internal capacity, risk tolerance)
- • A clear definition of what "failure" means
You leave the first session with fewer options, not more.
Stress-test assumptions and second-order effects
Most AI failures happen because assumptions that seemed safe turned out to be fragile.
I pressure-test:
- Vendor dependencies — What happens if they pivot, shut down, or raise prices? Do you have an exit path, or are you locked in?
- Organizational capacity — Can your team actually absorb this change, or are you assuming capacity that doesn't exist?
- Governance gaps — Who owns this decision after I leave? If the answer is unclear, this will drift.
- Second-order costs — What breaks downstream if this goes wrong? (Tool sprawl, team misalignment, vendor lock-in, training gaps)
- Technical feasibility — Can the vendor actually deliver what they're claiming, or is this marketing?
This is where "urgent" decisions often become "not now" decisions. If the foundation is weak, building on it is waste.
Commit to a decision sequence and governance
Once the decision is clear and stress-tested, I help you commit to a path that holds under pressure.
This includes:
- • A decision sequence — what happens first, second, third (no parallelization until the foundation is stable)
- • Governance structure — who owns what, how you prevent drift, how you handle new information without abandoning the plan
- • A "not now" list — what you explicitly will not do (this is as important as the roadmap)
- • Vendor management framework (if applicable) — how you evaluate, contract with, and oversee external partners
The goal is not consensus. The goal is a decision the organization can execute without fragmenting into tool chaos.
What you should expect in the room
I'm direct. I start by narrowing the decision, not expanding it. I force tradeoffs into the open.
You should expect:
- • Questions that challenge assumptions — not to be difficult, but to find fragility before you commit capital
- • Controlled pace — urgency is not the same as importance. I separate the two.
- • Fewer options at the end than at the beginning — more options = more risk of drift and misalignment
- • Honest feedback — if I think the timing or direction is wrong, I'll say it
- • No vendor bias — I'm not paid by tool companies. If a vendor solution is the right fit, I'll recommend it. If it's not, I'll say that too.
I'm not here to be liked. I'm here to make sure you don't waste money on AI theater.
Requirements from you
To work together, you need:
- ✅ Access to the decision-maker — I need to work with whoever has final ownership of the call (CEO, COO, CFO, or executive with budget authority)
- ✅ Willingness to hear "no" or "not now" — If you only want validation for a decision already made, don't reach out
- ✅ Readiness to document current reality — Not aspirational process maps, but how work actually happens today
- ✅ Commitment to sequence decisions — No tool shopping until the strategy is clear
- ✅ Honesty about constraints — Budget, timeline, team capacity, risk tolerance. If these aren't clear, we'll clarify them first.
If these are missing, this becomes theater. I don't do theater.
What success looks like
Success is not a document. It's the outcome.
After working together, you should have:
- ✅ Fewer active directions and stronger alignment — everyone knows what we're doing and why
- ✅ A clear "not now" list — what you're explicitly not doing (prevents drift and scope creep)
- ✅ Decisions that remain stable when new information arrives — the plan survives contact with reality
- ✅ A team that can execute without me — that's the goal. If you still need me after 6 months, something went wrong.
- ✅ Vendor relationships that are managed, not reversed — you control the relationship, not the vendor
The point is to make you independent, not dependent.