George Hanlon

Why I’m building Modularity

What I’ve seen up close is this: the people doing the most important client work spend a staggering amount of their week assembling documents. Suitability reports, onboarding packs, engagement letters, KYC checks. The thinking is the easy part — it’s the assembly, the formatting, the consistency-checking that eats the hours.

Generic AI tools don’t solve this. They aren’t built for regulated workflows, they don’t understand what a compliant document actually needs to contain, and no principal wants client data going somewhere they can’t account for.

Modularity is my answer: tools for modular workflows, for building AI tools and business applications that are genuinely applicable and useful. Purpose-built, not general-purpose.

Why modularity

The name is the thesis. For narrow, repetitive business tasks, a small purpose-built tool beats a general-purpose model on cost, reliability and governance. A scoped tool runs a fraction of the compute a general model burns on the same work — you pay for the task, not the platform.

And modules recombine. The same building blocks reassemble into different tools for different processes, so every new application is cheaper and faster to build than the last. Each tool carries its own process, can be tested against the one job it does, and stays small enough to audit. That, I believe, is the key to AI that’s affordable to run — and to business applications that actually work.

George
Founder, Modularity
[email protected]