How to run public-record due diligence on a company

The sources, the signals, and how to collapse a six-hour records search into one query.

2026-08-20 · 6 min read

Due diligence on a company is, at its core, a records problem. Before you sign, lend, buy, or onboard, you want to know what the public record says: is the company real and in good standing, what has been filed against it, and what risks are hiding in systems you would never think to check. The hard part has never been reading one record. It is that the records live in a dozen separate government systems, each with its own search box and its own format.

This guide walks through what public-record due diligence actually covers, the signals that matter most, and how to collapse the whole search into a single query.

The sources public-record diligence pulls from

A thorough public-record check spans federal, state, and local systems. At the federal level: SEC filings, SAM.gov registration and exclusions, SBA data, EPA environmental records, and federal court dockets. At the state level: Secretary of State business registrations, UCC financing statements and liens, and professional and business licenses. At the local level: county property and deed records, and building permits.

Each of these answers a different question, and each normally requires a separate visit to a separate portal. That fragmentation is the reason diligence takes hours.

The signals that actually matter

Not every record is a red flag, but a handful of signals do most of the work. Corporate status and registration tell you the entity is real and active. UCC liens and financing statements reveal what the company has borrowed against. Litigation and judgments show disputes and liabilities. Bankruptcy is an obvious flag. Regulatory and environmental actions surface compliance risk. Government exclusions and debarments tell you whether the company is barred from federal work.

Read together, and attached to the right entity, these signals answer the real question: not whether a company exists, but what the public-record risk profile of this business is.

Why entity resolution is the hard part

The catch is that a company rarely appears the same way twice. It registers in multiple states under slightly different names, holds assets under subsidiary LLCs, and gets named as free text in court records. If you match on names alone, you either miss records that belong to the company or wrongly attach records that do not.

Reliable diligence depends on entity resolution: joining records across sources on the strong identifiers that survive between them (registration numbers, UEI, EIN), so one company is one entity even when it appears in sixteen datasets. A lien filed under a subsidiary still surfaces on the parent.

How to do it in one call

Amalgament resolves those sources into one entity and returns the whole public-record picture in a single query. Instead of visiting a dozen portals, you ask once, over a REST API or an MCP connection, and get the registrations, filings, liens, litigation, sanctions, and property already joined to the right company. Because the same layer is agent-native, an AI diligence tool can assemble the report mid-workflow.

The result is that a public-record due-diligence report goes from an afternoon to an API call, and every record keeps its source and date so the report is auditable.

See the due-diligence solution

Common questions

Is public-record due diligence the same as a background check?

No. Public-record business diligence is about companies, not consumers. It is non-FCRA and covers business entities, so it is used for KYB, vendor vetting, and deal diligence rather than decisions about individuals.

What is the fastest way to run it?

Query a layer that has already joined the sources and resolved the entity, so the whole record comes back in one call instead of a portal-by-portal search.

Due diligence · What is KYB · Entity resolution API

Keep reading: Permit-based lead generation: finding businesses about to spend · KYB vs KYC: what business verification actually checks · How entity resolution works (and why name-matching fails) · What public records reveal about a commercial property · How to give an AI agent access to public records (MCP)