Real-World Impact

Most enterprise AI initiatives stall in the gap between a convincing demo and a deployed business outcome. The technology is rarely the constraint. The operating model is.

After 25 years selling transformation into large enterprises, the pattern holds across every cycle, from ERP to cloud to AI: adoption follows value quantification, owned at board level, expressed in the language of the buyer's industry. A generic AI pitch dies in procurement. An industry-quantified one reaches the board, because it answers the only questions a P&L owner actually asks: what changes in my numbers, by when, and who is accountable.

Becoming AI-native is therefore a commercial and organizational discipline at least as much as a technical one. It means rethinking how decisions are made, not just which tools are bought. It means treating data quality and core systems as the substrate that determines whether AI produces measurable outcomes or expensive pilots. And it means equipping the people who own revenue and operations to carry the case for change themselves, rather than delegating it to an innovation function on the side of the org chart.

The companies that win this shift run it as a value program with an industry thesis, executive ownership, and quantified outcomes per use case. The companies that lose it run it as a technology rollout and wait for transformation to emerge. It will not.

This is the work I do: defining what AI-native means for a specific industry, building the quantified case at board level, and constructing the go-to-market that turns a frontier capability into adopted reality.

Value selling is widely claimed and rarely practiced. Practiced, it means engineering the case for change: quantify the executive outcome, anchor it in the economics of the buyer's industry, then make the motion repeatable so it survives beyond the people who invented it.

I was part of the early team that built the value engineering discipline at SAP, and I have applied it across 25 industries and three continents: to global groups, to mid-size manufacturers, and to a greenfield scale-up that almost nobody took seriously at the time. The discipline travels because it is not about the vendor's product. It is about the buyer's P&L.

It also travels to domains where ROI is considered unprovable. Sustainability is the test case: positioned as compliance, it is a cost center and a hard sell; positioned through value, linking ESG metrics to financial outcomes, it becomes a board conversation about growth and cost. The method is identical. Only the courage to quantify changes.

In the AI era this discipline matters more, not less. The technology has never been easier to demo and harder to justify. Buyers are drowning in capability claims and starving for quantified cases. The sellers who can engineer value, not just present it, will own the next decade of enterprise technology.

I invest and advise where AI meets the physical world: world foundation models, physical AI, agentic systems. These are not incremental improvements to enterprise software. They are the next industrial layer, and they will be adopted, eventually, by the same enterprises I have spent my career selling transformation into.

That is my edge as an investor, and I am clear-eyed about what it is and is not. It is not technical depth for its own sake. It is the adoption lens: 25 years of quantified, board-level selling taught me the real distance between a frontier capability and a P&L, and what it takes to close that distance, industry by industry. Most frontier companies fail commercially not because the science is wrong but because that distance was never mapped.

The relationship runs both ways. Backing frontier companies keeps my enterprise view honest about what is actually coming, years before it shows up in an RFP. The enterprise view keeps my investing honest about what will actually land, and when.

I also build with my own hands. Founding AI ventures and working directly with agentic frameworks is how I keep my judgment grounded in what the technology can really do today, not what a keynote says it will do. At the frontier, fluency is not optional.

Scaling a practice or a go-to-market motion is not a hiring problem or a content problem. It is a transfer problem: how do you move judgment, not just material, from the founding team to people who were never in the room?

The answer I have applied across three continents has three parts. Codify the judgment: turn what the founding team does instinctively into a methodology that holds without them. Enable at scale: train the field not on slides but on the motion itself, until thousands of sales professionals carry a conversation they did not invent. Build the community that sustains it: peers correcting and improving the motion long after the program ends.

The same logic applies across cultures, which is where most scaling efforts quietly fail. I have built and led teams from France to Japan, Korea, India, and Southeast Asia. The methodology that transfers is the one expressed in outcomes and economics, because outcomes translate. Style does not.

What breaks organizations that scale content instead of judgment is predictable: the material spreads, the conversations flatten, and quality decays with distance from the founders. What scales judgment is harder to build and much harder to copy. That is the point.