Technology for institutions that answer to the public.
Strategy, software, data and applied AI, built for an island with two jurisdictions, two languages and two currencies.
We build and operate our own.
Both are pre-launch and neither is in public use yet. They are shown as evidence of what we build, not as a customer list.
What we do.
Technology strategy
What to build, what to buy, and what not to do. Written so that a board can act on it and a procurement officer can evaluate it.
Software and applications
Custom systems designed, built and maintained. Web and mobile, built to run on the connectivity people actually have.
Data and registries
Records, registries and the systems that hold them. Cadastral, statutory and operational data, with provenance carried through to the interface.
Applied AI
Retrieval, attribution and classification over large document sets, with the evaluation suite that proves it works. Used where it can be evidenced, and not where it cannot.
How we work.
Every figure carries its source
A number without a source and an observation time is not evidence. Our interfaces show both, because in public work the provenance is the product.
Gaps are stated, not filled
Where the data is incomplete, the system says so. Filling a gap with a plausible estimate is how institutions lose the ability to tell what they know from what they assume.
Coverage is documented before claims are made
Every dataset we build ships with a written statement of what it does not contain. It is the first document we write and the last one we revise.
Built for the conditions here
Two jurisdictions, two languages, two currencies, and patchy connectivity. Systems that assume otherwise fail quietly and expensively.
The people behind the work.
Work directly with the principal, from conversation to delivery.
Our work spans strategy, software, data and applied AI. Our ventures, 1648 MLS and SXM Today, are in development.
Based in Philipsburg. Named for 1648, built for both sides of the island.
Built to WCAG 2.1 AA for public-sector work.
Tell us what you are trying to do.
Tell us about your project.
What are you trying to achieve, and where are you getting stuck?