What I do
Seven domains. One conviction.
I have never been interested in technology for its own sake. These seven are simply where I have worked, and in each of them I was asking the same two questions. Does this change anything for the person at the other end? And can the company still do it six months later, when the launch is over and nobody is paying attention?
The domains changed as I moved: drives and power electronics, then buildings that sense themselves, then software, then electrification, then companies of my own. The method did not. Understand the physics or the economics well enough to be specific about them. Surface the trade-offs early, while changing your mind is still cheap. Then build the cadence that holds the promise after the launch. That last part is where most of it is won or lost, and it gets the least attention.
Energy, Electrification & Power Systems
Eleven years in power electronics, then the operating seat of an electrification business.
I joined Danfoss in 2007 and stayed eleven years, moving through application engineering, project management, R&D, and finally global responsibility for market quality on the drive product lines. The application years shaped everything after them. I gave global expert support on electric and permanent-magnet motor applications across HVAC, water and wastewater, automation, and food and beverage plants, which meant that when a drive and a motor would not behave in a real installation the way the datasheet promised, the question arrived on my desk. A specification and an installation are two different documents, and I learned that at the age where it still changes how you work. Years later I took the operating seat at an electrification and battery storage company, where converting a diesel fleet to electric had almost nothing to do with the motor. It was parts arriving on time and vehicles going out on time. It was serviceability, R&D, the ISO certification, and getting the telemetry right. And more than any of that, it was building the processes that did not exist yet, reaching a new level of productivity, and getting engineering, production and service to cooperate as one organisation instead of three.
Buildings, IoT & Connected Products
Sensors, controls and connected products, owned as one system.
Hardware, firmware, software and cloud usually sit in four teams with four roadmaps, and the seams between them are where connected-product programmes quietly die. I owned R&D strategy for IoT-based HVAC innovation in a global organisation of forty-odd people, and later rebuilt an R&D function that had just been cut, then shipped a connected sensing platform out of it. One thing from that period I still use. When the physics, the interface and the service contract disagree, the product goes late, and the lateness keeps looking like a software problem long after it has stopped being one.
Software, Product & Platform Engineering
How software actually gets built, and kept alive afterwards.
Architecture, cloud and infrastructure, pipelines, supply-chain security, and the dull machinery that decides whether a team ships every week or every quarter. I have run engineering organisations, and I still write code in my own companies. Mostly that is because it is very hard to argue with me about an estimate for something I shipped myself last month.
Strategy, Commercial & Growth
The plan, and the pipeline it has to survive.
Value proposition, pricing, key accounts, partnerships, and the marketing engine underneath all of it. I have carried the number as well as written the strategy, and the two jobs teach you different things. Pricing I treat as product management, because that is what it is. Marketing automation I treat as operations, because the day after it goes live somebody has to own it.
Operating Discipline & Turnaround
P&L, cadence, governance, and the quarters that go wrong.
Full P&L for an international business unit, execution cadence, governance and quality systems, including the part of the range nobody puts on a website, where orders slip and the funding does not arrive. I have rarely met an organisation that lacked talent. I have met plenty that had no rhythm, so the trade-offs only became visible at the point where they had already become expensive.
Venture Building
Six companies running, one exit.
Commerce for local shops. Booking for salons. A marketplace for language teachers. Fintech tooling for brokers. Trading education in two languages. Every one of them started because I could name the problem precisely, usually because it had annoyed me or somebody close to me first. They are also where I find out what works before I write about it.
Fintech, Crypto Rails & Trading Systems
Payment rails, broker infrastructure, and systems that trade.
Broker and settlement tooling, stablecoin payment rails for markets where cards are simply not available, automated trading engines, and machine-learning models that re-fit themselves as the regime changes. I run all of it with my own capital and I publish the track records. That is an uncomfortable standard to hold yourself to, and I have kept it anyway.