About

I have spent a decade in engineering leadership, most of it in the part of a company's life where the team is growing faster than its decisions are.

I now work as a fractional CTO, deliberately with two clients at a time. That limit is the whole model: it is enough to be genuinely inside a company, and few enough that I can still be the person who is called at the wrong hour.

How I work

• Ownership before architecture Most architecture arguments are ownership arguments in disguise. Write down who owns what, and half the diagram settles itself.
• Measure before you rewrite One review in three cancels a planned rewrite. The deck is usually well argued and usually wrong about where the time goes.
• Reversible decisions, made quickly Spend the deliberation budget on the handful of bets that are expensive to undo, and make everything else fast and reversible.
• Leave the decision behind, not the dependency If the team cannot make the call after I leave, the engagement failed, however good the quarter looked.
• Write the bar down before you hire to it A hiring problem is usually an undefined bar. Put the bar in writing, build the loop that tests for it, and hand the loop to your own engineers to run.