Things we write about,
things we open-source,
things we think are true.
Product deep-dives, compliance research, open-source tooling, and our take on where this industry should be headed. Written by the people building Mercurium.
The FinCEN Files and why RegTech still cannot catch them
$2 trillion in suspicious transactions flagged by the largest banks — and processed anyway. We break down why existing RegTech architectures structurally cannot solve this, and what a platform built for the job would actually look like.
Why we rebuilt document intelligence from scratch
Every RegTech platform claims to "read" documents. Almost none of them actually do. Here's what we learned after parsing 600-page prospectuses — and why we had to build our own pipeline from the ground up.
SRA and FCA AML onboarding requirements for solicitors
With AML supervision for legal services moving from the SRA to the FCA, this note clarifies what is changing, what is not, and how solicitor firms can prepare.
Research & Open Source
Onboarding Gap: Why Client Onboarding Still Fails at the Edges
Industry and regulator data points to a consistent onboarding gap. Across financial services, 30% of client journeys are still not cleanly handled by existing tools. They abandon, stall, escalate, require manual review, or create material compliance rework. This note asks what is failing, and why?
mercurium-document-format: our open-source layout spec
We're publishing the canonical document schema every Mercurium agent consumes — hierarchies, definitions, cross-references, annotations — as an open specification. One format, any document, any downstream model.
Benchmarking UBO chain resolution across 80+ jurisdictions
We evaluated every major corporate-registry source against a 2,400-entity test set spanning 80 jurisdictions, trusts, SPVs and nominee arrangements. Here's what we learned, where the holes are, and what a credible UBO layer looks like.