Blog
-
4 Oct 2026
Stop starting every order from a blank page
Why repeat work deserves a template the business controls itself, and the two design calls that made it safe to rely on.
-
7 Sep 2026
The first system built on our shared toolkit found four things wrong with it
What actually reusing our internal packages, instead of just intending to, taught us about them.
-
26 Aug 2026
Making invisible errors visible before the client has to tell you
What changed when a routine bug in our own error logging turned into a decision to stop treating “no news” as “no problems.”
-
24 Aug 2026
The contract system that flags its own mistakes before a client ever sees them
How we replaced a five-person manual sign-off chain and a hand-edited legal document with one system that assembles the right contract, checks its own cross-references, and always knows exactly where…
-
19 Aug 2026
The checklist that gets smarter every time it’s used
A repeatable enquiry process doesn't have to start from zero every time. Here's what changes once past answers become something your team can search, instead of something someone has to…
-
13 Aug 2026
Proving a premium service tier is actually faster
How we replaced a marketing promise with a number a manager can check on any given day.
-
10 Aug 2026
From “Can Someone Pull Those Numbers?” to a Report That’s Just There
How we replaced four scattered, manually-answered management questions with reports built on one shared foundation, instead of four separate one-off fixes.
-
5 Aug 2026
Stop Shipping Code Just to Change What Your AI Writes
A practical look at moving an AI content prompt out of application code and into the hands of the team that actually knows if the writing is good.