A working AI foundation in four to eight weeks.
Fixed scope, known duration, defined outcome. You end with governance your counsel can read, tooling that is actually stood up and configured, and people trained to use it on their own work.
Piloting and integrated are not the same business.
- Revenue gap4×
Organizations with fully integrated AI are nearly four times more likely to report revenue growth than those still piloting — 58 percent against 15 percent.
Source · Grant Thornton 2026 AI Impact Survey - Scale gap7%
Report fully scaled AI, and 39 percent report enterprise-level EBIT impact — against 88 percent who use AI somewhere in the business.
Source · McKinsey Global Survey on the State of AI, November 2025 - Governance gap36%
Have a formal AI governance framework. Three quarters have an AI usage policy — so the policy exists, but the framework underneath it usually does not.
Source · Pacific AI 2025 AI Governance Survey - Agent readiness1 in 5
Companies have a mature governance model for autonomous AI agents.
Source · Deloitte State of AI in the Enterprise 2026
The distance between those two positions is rarely the model you picked. It is governance nobody wrote, licenses nobody configured, and staff who have never been shown what good use looks like in their own job. That is what this program builds, and it is why it has an end date.
Four phases. The shape is fixed; the duration scales with scope.
Every engagement runs the same four phases in the same order. What changes is how long each takes — driven by how many workstreams are in scope and how many people need training. Nothing is discovered halfway through and added to the invoice.
Assess
Four layers get scored rather than discussed: strategy and culture, data and governance, technology, and people and skills. Every proposed AI use is mapped to a risk tier, and the regulatory picture for your sector and jurisdictions is inventoried. Self-assessment usually runs ahead of the evidence, and this is where the difference gets settled — before anyone spends money.
Decide
A facilitated executive session that ends with a decision, not a brainstorm. Candidate domains are scored against six criteria, processes within the chosen domain are placed on a type-by-risk grid, and the top candidates are scored again. Choosing one is the discipline that pays: only 4 percent of companies take a focused, single-domain approach, and those that do earn twice the ROI (2024 BCG survey of 1,000 CXOs, reported in Harvard Business Review, November 2025). You leave with a written executive AI vision, a prioritized use-case list and a named accountable sponsor.
Establish
The governance framework, acceptable-use policy and AI standards are drafted against your risk tiers and mapped to NIST AI RMF. In parallel, tooling recommendations are made and then acted on — licenses stood up, tenancy configured, access policy set. Guardrails that fit on a page beat a manual nobody opens.
Enable and hand over
Training sessions run against your prioritized use cases, not generic demos, so people practice on work they actually do. The engagement closes with an AI roadmap, a 90-day action plan with a named owner per item, and an executive readout. Everything produced is yours.
Three workstreams. You choose how many you need.
The program is not a document pack. It is three workstreams, each producing artifacts you keep and operate. Take one, take all three — that choice is what sets both the duration and the price.
- T&R
Training and readiness
The irreducible core: bringing people to the point where they can actually use AI on their own work. Executive vision first, then use-case discovery and prioritization with the people who do the work, then training built on those use cases. Taken on its own, this is the four-week engagement.
Executive AI vision · use-case discovery · use-case prioritization · training sessions · executive readout
- GOV
Governance
Create the governance you are missing. Risk tiers for every category of AI use, a small set of clear boundaries rather than a thick manual, and a review path proportionate to risk. The framework maps to NIST AI RMF, so your counsel, customers and insurers have a recognized standard to check it against.
AI governance framework (maps to NIST AI RMF) · acceptable-use policy · AI standards
- LIC
Licensing
Advice on tooling, then the work of putting it in place. Architecture recommendations are grounded in the prioritized use cases rather than a vendor shortlist, and the licenses are then established and configured — tenancy set up, access policy applied, admin handed to your team.
AI architecture recommendations · licenses established and configured · access policy
- ALL
Produced by the engagement as a whole
Three artifacts come out of the program regardless of which workstreams are in scope, because they describe where you are, where you are going and what happens first.
AI maturity assessment · AI roadmap · 90-day action plan
Two configurations: training and readiness alone is the four-week engagement; all three workstreams together is the eight-week engagement.
What you can do the week after we finish.
These are capabilities, not documents. Each one is something the organization could not do before the engagement started. Which of them you end up with depends on the workstreams you take.
Decide on AI requests in days
Every request has a risk tier and a written standard to check it against, so routine, low-risk uses stop queueing behind an executive committee and genuinely consequential ones get real review.
Put approved tools in people’s hands
Licenses are live, access policy is applied, and staff have been trained against use cases from their own work rather than a vendor demo. Adoption starts from something they have already practiced.
Answer the governance question
When counsel, a customer, an auditor or an insurer asks how you govern AI, there is a framework mapped to NIST AI RMF, an acceptable-use policy and a set of standards to hand them.
Buy tooling on evidence
Architecture recommendations trace back to prioritized use cases and your risk tiers, which makes the next licensing conversation a scoping exercise rather than a vendor bake-off.
Know what happens first
A 90-day action plan with a named owner per item, sitting under a roadmap. The program ends with the next quarter already decided, not with a recommendation to decide it.
Most AI programs stall on operating-model decisions made early and badly, not on the technology. That is a fixable problem, and it has an end date.
- 30 yrtechnology leadership
- Academic medicineEmory School of Medicine · Stanford Medicine
- EnterpriseHP · Cisco · Borland
- MIT xProAI for Senior Executives, completed July 2026
Two configurations, priced before it starts.
The starting price buys the four-week engagement: training and readiness, plus the artifacts the program produces in every configuration. Adding governance and licensing takes it to roughly eight weeks and is quoted at scoping. Whichever you choose, scope, duration and price are agreed in writing before the work begins.
Starts at $15,000
Four weeks, training and readiness. Taking all three workstreams runs to roughly eight weeks and is quoted separately.
- Every configuration · AI maturity assessment
- Every configuration · AI roadmap and 90-day action plan
- Every configuration · Executive readout
- Training and readiness · Executive AI vision
- Training and readiness · Use-case discovery and prioritization
- Training and readiness · Training sessions for your teams
- Governance workstream · AI governance framework, mapped to NIST AI RMF
- Governance workstream · Acceptable-use policy and AI standards
- Licensing workstream · AI architecture recommendations
- Licensing workstream · Licenses established and configured
The things buyers ask before signing.
Build Your AI Foundation.
A short scoping conversation sets the workstreams, the number of people and the calendar. You get a fixed scope, a fixed duration and a fixed price before anything starts.