Mandates keep arriving and each one assumes you already have inventoried, described, semantically defined data. TopQuadrant builds that foundation on open W3C standards, aligned to NIEM and BFO, owned at the domain level so it survives reorganisation and the next administration’s priorities.
Mandates that assume capability you do not yet have, standards proliferating without a translation layer, and backlogs that trace back to undescribed data.
Federal AI direction, the FDTA joint standards, the Evidence Act, and the OPEN Government Data Act all assume inventoried, quality-controlled, semantically defined data. The mandates arrive with deadlines and without the underlying capability.
DCAT-US, NIEM, BFO, USCDI, and the FDTA taxonomies all coexist, and nobody owns reconciling them. Every cross-agency initiative rebuilds the same mapping work from scratch.
Directives to eliminate information silos and Privacy Act obligations are both enforceable, and the tension is now litigated rather than theoretical. Both positions have to be defensible from the same data estate.
Executive branch agencies closed FY2025 with 463,541 FOIA requests pending. Behind the processing capacity sits the same root cause: records that are not described, not linked, and searchable only by full text.
When a domain model is governed as infrastructure, the next programme starts from it instead of starting over.
The use cases financial data, risk, and governance teams run on TopBraid EDG.
Open W3C standards underneath, and the industry’s own vocabularies on top, so nothing you model is locked to us.
Browse the resource hub →NIEM
Model v6.0 became an OASIS Standard in December 2025. Align governed models to NIEM domains without translation.
BFO
The ISO/IEC 21838 top-level ontology, now formally collaborating with NIEM, supported natively in OWL.
DCAT-US
Describe assets once and publish conformant inventories, so a schema change is a mapping update.
W3C
The stack NIEM, BFO, DCAT-US, and the FDTA standards all ultimately rest on, used directly rather than approximated.
Each one needs a governed answer that carries its definition and its lineage, not a best guess from a warehouse query.
“The model outlived the programme that built it. That has not happened before here.”
Guides, sessions, and field notes for financial services data teams.
View all →Reading the recent BCBS 239 findings as a data architecture brief.
WebinarWhy constrained query construction beats a better prompt.
BlogWhere the standard helps, and where you still model your own.