About

Twenty years between engineering and the organisations that commission it, mostly in the parts where the two do not understand each other yet.

How I got here

I started in enterprise integration, which is a polite way of saying I spent five years discovering how public-sector systems fail at their seams. That led to a doctorate on coordination failure in distributed grantmaking — the same problem, different vocabulary — and then to running data and platform for a community foundation, where I had to live with the consequences of my own architecture decisions for four years. I recommend it. Nothing improves a design instinct like maintaining the thing.

How I work

Three habits, and they are the whole method. I write down what I decided and why, next to the thing it explains, because the reasoning is worth more than the decision. I try to say plainly when something is a guess, since a confident guess is worse than an admitted one. And I would rather hand back a smaller recommendation I can defend than a larger one I cannot.

What I am usually brought in for is a decision that has been deferred: whether to rebuild or repair, whether a model is doing real work, whether a reporting burden is the funder's fault or the system's. These are rarely technical questions wearing technical clothes — but they are usually answerable, and answering them is most of the value.

Elsewhere

There is a CV, a list ofpublications, and onerecorded talk about a result that did not go the way we wanted.