I solve the problems that sit between what a customer needs and what technology can actually do.

The product is rarely the answer on its own. The job is to understand the workflow, the people, the constraints and the value at stake, then work backwards into a solution that changes the outcome.

Across more than three decades, the pattern is remarkably consistent: learn the system deeply, understand the operating problem, help shape what the solution needs to become, explain an emerging idea to the market, and prove that customers will act on it.

I used the period after BlackLine to write, build and test rather than simply study AI. Some of the outputs are businesses, some are products, and some are tools I built because I needed them. The common thread is learning by putting AI into real workflows.

The new capability is hands-on AI solution development. The older capability is still the same: understand the problem well enough to know what should be built.

Models are components, not the architecture. I design around the capability, choose models for the job they are doing, and keep enough separation that the underlying technology can change as the market moves.

The book is another example of the same process: see a recurring problem in the field, understand the pattern, codify a point of view, then turn the thinking into something people can actually use.

People who managed me, reported to me and worked beside me repeatedly describe the same things: whole-picture thinking, customer focus, creativity and the ability to translate complexity into a useful solution.