What we do
From scattered records to a resolved graph.
Apriori does three things: we acquire authoritative public data at the source, we resolve it with AI into a single graph of organizations and licensed professionals, and we deliver it with provenance intact. This page explains each step.
1. Acquire at the source
We collect US public records spanning regulatory filings, disclosure documents, registrations, licenses, and government spending - federal, state, and county - and international business registries, directly from the systems of record. That means thousands of file formats and dozens of languages, handled by pipelines built for exactly that. Transliteration and multi-language processing are core capabilities, not afterthoughts.
2. Resolve with AI into the organization graph
Entity resolution is our specialty. LLM-assisted pipelines link, deduplicate, and reconcile records across sources into one resolved view per organization - its registrations, its filings, its licenses, its awards - and per licensed professional. The models propose, deterministic rules and provenance checks decide, so results are reproducible. The outcome is a maintained graph of US and international organizations spanning commercial, nonprofit, and public-sector organizations - each entity linked to the government records behind it.
3. Deliver with provenance
Three delivery modes: bulk datasets, API access, and monitoring - change detection on new filings, status changes, and registration events. Every attribute stays traceable to the government source it came from, whatever the mode.
Where it's used
Built for regulated work.
Identity verification & KYC/KYB
Verify organizations and the licensed professionals behind them against the official record, with citations your auditors can follow.
Compliance & due diligence
Officer and director history, entity status, and filing trails - resolved across jurisdictions.
Risk & monitoring
Clean, current entity data as the substrate for risk models - with change detection when the record changes.
Data infrastructure for AI
Deterministic, deduplicated, source-cited data that RAG and LLM products can quote without hallucinating.