Between us, we have spent the past two decades inside technology firms — building and selling them, co-founding and running cloud consultancies through to acquisition, and shipping AI engineering into production at the moments where the field was being figured out, not described. We have worked together for most of the last ten years, in the same room on the same firm — long enough to know what the others will say, short enough that we still argue about it.
From those seats, we watched the same conversation play out at hundreds of leadership tables. They had agreed AI was real. They were trying to figure out what to do about it. They had a Director of AI, a steering committee, two hundred pilots, three in production. They had a vendor's deck that ended with the words scale next quarter. Very little of it was compounding.
What was missing — almost everywhere — was a firm that would sit on the inside with them. Not run pilots. Not author roadmaps. Do the work that gets a system into production, then stay long enough for the system to compound. That firm exists in pieces — at the hyperscalers, at the frontier model providers, at the better consultancies, scattered across the alumni networks of all three. We started Revoleap to gather those pieces, and to do that work for a small portfolio of consequential firms each year.
We do five things — AI Engineering, Cloud Engineering, Data Engineering, App Engineering, and Platform Operations — and we do them as a single piece of work. The five practices are how we organise the firm internally. They are not how we organise the engagement. The engagement is one team, with a named partner accountable, end-to-end.
If you are at the moment in this letter — past "AI is real" and into "what now?" — we should talk.
