Baselayer

Machine learning clusters identity anomalies, spots velocity patterns and fraud rings across the network and drives agents such as Risk Co.Pilot. Underneath sits data aggregation and deterministic matching over public filings, a business verification service that would sell on its own.

Low risk applications pass automatically, while high risk cases reach a person with explainable risk signals showing why. Review queue design, threshold settings and sampling of automated approvals are not described.

Match rates and auto approval rates measure how often the system says yes, not how often an approved business proves fraudulent. Decisions are called explainable, but accuracy figures, model documentation and an evaluation method are unpublished.

Nuvei raised match rates to 98 percent from about 60, LoanPro cut manual reviews by 60 percent, and Breakout Finance reports 98 percent lien matching accuracy. The figures come through the vendor with no baseline period.

Onward training is ruled out in writing: no platform may use Baselayer data to train its models, and activity linked across institutions is anonymized. Inquiry patterns still reveal which businesses an institution is evaluating, and what anonymized means in practice, whether participation is optional and how long contributed data stays are all undefined.

Officers, beneficial owners and signers are covered directly in a privacy policy issued by Osiris Ratings, Inc., doing business as Baselayer, which lists what is kept about the people connected to a looked up business. Retention lasts only as long as records stay useful, and no hosting provider is named.

SOC 2 Type II compliance under the AICPA framework is stated on the homepage. Trust portal, report request route and audit period are not offered.

A plain position on the Fair Credit Reporting Act appears in the privacy policy: Baselayer is not a consumer reporting agency. Military Lending Act and death master file checks sit on the pricing page, and the scope maps to customer identification and beneficial ownership duties for business accounts.

Online presence feeds credit risk scoring, so websites, social media and reviews shape a business's access to capital. Thin online footprints track owner age, rural location and immigrant ownership more than creditworthiness, and no demographic analysis or error rate is published.

Flags such as synthetic or high risk can cost a business an account or loan, and network signals follow it to the next institution it applies to. Baselayer publishes nothing on who bears a wrong judgment and describes no audit trail, flag reason or route for the business to contest, though decisions are called explainable.

Sources are named in the privacy policy: secretary of state filings, court and lien records and sanctions lists, combined with proprietary and partner data. Providers for the clustering, scoring and agent components are undisclosed.

Delivery runs through a public API documented at docs.baselayer.com, an embeddable console and an MCP server for agents, with Exa and Nevermined as agent partners. Embedding in FIS Decision Solutions and ChexSystems reaches institutions through a core processor's portfolio, though it names no core banking or origination system integration.

Application and inquiry records from thousands of institutions sit in one shared layer. Hosting provider, region and data location are undisclosed for this domestic cloud service.

Plans are priced openly: Startup from $150 a month for 500 credits and Growth from $750 for 3,000, with no contract and a quoted Enterprise tier. Credits consumed per check are not listed, so a rate per verification cannot be worked out.

Banks and credit unions, fintechs and small business lenders head a segment list that also reaches marketplaces and the agentic economy. Coverage is domestic, verifying American entities only, with insurers and capital markets unaddressed.