Enigma Technologies

Proprietary algorithms combine thousands of sources to resolve business identity and derive financial health signals, and a know your business compliance agent shipped in 2026, so models do meaningful work in entity resolution and classification.

What holds this at B is that the decisive asset is a data supply position rather than a modelling one: a panel covering more than 40 percent of United States card transactions is an access advantage no algorithm confers, and independent analysis describes the company as shifting from descriptive to predictive analytics, which indicates the current product is largely the former.

The historical product supplies data for a lender's own models and analysts to use, which places all judgement with the institution, and the newer know your business compliance agent moves in the other direction by performing verification work itself. Nothing describes what that agent may conclude on its own, what confidence accompanies an identity match, whether a failed verification routes to a person, or how an institution configures the threshold. For a component feeding origination decisions on small businesses, the absence of any stated boundary is the gap.

No accuracy figure, match rate or validation result was located, and the claim to be the most accurate and comprehensive source is asserted rather than demonstrated. The panel share is a coverage statistic rather than a correctness one, and the two diverge precisely where it matters, since a business partially captured by the panel yields a revenue figure that is verifiably derived and still incomplete. Reported outcomes describe effort saved and losses avoided rather than measured precision.

A prominent corporate card and spend management company is named as having built its fraud intelligence pipeline on this data, which is a specific technical dependency rather than a logo. Distribution partnerships are named and dated, including placement on a major data platform's marketplace using zero copy sharing, with that platform's technology partnerships director quoted, and an integration with a leading enrichment orchestration tool.

Total funding exceeds 200 million dollars from investors including a large venture firm and a major insurer. Coverage is quantified at more than 49 million United States businesses with financial health intelligence on over 33 million, resting on a panel exceeding 40 percent of national card transactions.

Privacy preserving computation addresses the central concern structurally, since the point of the technique is that parties contributing or consuming data do not see each other's underlying records, and the transaction panel is aggregated and anonymised before it reaches any customer. That is a stronger answer than most data businesses offer.

Held at B because the boundary question is only half addressed: nothing states whether customer queries, outcomes or feedback are retained and used to improve the models and scores served to other institutions, which for a shared intelligence layer is the remaining exposure.

The privacy architecture matches the sensitivity of what is held, which for a panel spanning more than 40 percent of national card transactions is a very high bar. Transaction data is described as anonymised before it becomes product, the company holds a Type II service organisation control attestation, and it has implemented privacy preserving computation, the same class of technique recorded at Omnisient, so analysis can occur without exposing underlying records.

Independent analysis notes this was done proactively and that it reduced procurement friction with regulated institutions, which indicates the controls were built to be examined rather than described. That combination of anonymisation, cryptographic technique and independent attestation is the most complete privacy position of any data business in this index.

The company holds a Type II service organisation control attestation, the version that tests whether controls operated effectively over a period rather than whether they were adequately designed at a point in time, and independent analysis records that obtaining it was a deliberate move to meet heightened data privacy expectations which then reduced procurement friction with regulated financial institutions.

For a business whose entire asset is a panel of anonymised card transactions covering a large share of a national economy, an independently attested control set is the minimum a bank should require and the overwhelming majority of vendors in this index still do not publish one.

No supervisor, statute or instrument is named, and independent analysis identifies regulatory limits on data use as one of the principal risks to the business, which is an acknowledgement rather than a disclosure. The question this product raises is specific: supplying financial information about businesses to lenders for credit decisions sits near the consumer reporting framework wherever small business assessment reaches the owner, and nothing published addresses where that line is drawn or how permitted use is controlled.

The inclusion claim is quantified and the mechanism behind it is genuinely fairness improving. Because revenue is derived from observed card transactions rather than modelled from employee counts or industry codes, a business is assessed on what it actually earns instead of on a proxy for its size, and proxies of that kind encode structural bias in a way direct observation does not.

The company states this has helped lenders identify hundreds of thousands of healthy small businesses that would otherwise have been overlooked or denied credit, which is an outcome rather than an aspiration. The counterpart is a coverage inversion: a business taking little card payment, because it is cash heavy, rural, invoice based or serving communities that transact differently, is invisible to the panel, so the businesses hardest to see remain hardest to fund. No coverage analysis by business type is published.

No guarantee, indemnity or falsifiable accuracy commitment was located. The institutional customer can test the data against its own portfolio outcomes over time, which is genuine recourse for a data product. The small business has none and is unusually exposed: its revenue, growth and transaction patterns are observed through the card panel and sold to lenders and issuers, it is never told, cannot see the figures attributed to it, and has no route to correct a revenue estimate that understates its business and costs it credit.

The primary source is disclosed with unusual specificity for a data business, stated as a panel of more than 40 percent of United States card transactions, supplemented by thousands of other sources, and the company draws a clear distinction between figures derived from observed transactions and estimates modelled from employee counts or industry codes. That tells a buyer exactly what kind of evidence sits behind a revenue figure.

What is not disclosed is the panel's own provenance, meaning which processors, issuers or intermediaries contribute the transaction flow, and that determines both the coverage profile and the permitted uses of everything built on it.

Distribution is designed around meeting decisions where they already happen. Data is published on a major cloud data platform's marketplace using zero copy sharing, so an institution can join it to its own tables inside its existing warehouse without moving anything, which removes the largest practical obstacle to adopting an external data source.

A separate integration with a leading enrichment orchestration tool reaches commercial teams, interface access serves real time verification, and warehouse integrations handle bulk. Three named routes covering analytical, operational and orchestrated consumption is unusually complete for a data provider.

No hosting provider, region selection or residency commitment was located from the company. Marketplace distribution through zero copy sharing means much of the analytical consumption occurs inside the customer's own environment, which reduces the question's weight without answering it, and nothing describes where the underlying panel is processed and held.

No pricing, packaging or basis of charge was located. Availability through a major data platform marketplace provides a procurement route that lets customers consume the data under existing commercial arrangements, which lowers friction without disclosing terms. Nothing indicates whether charge falls per business record, per query, by data volume or as a licence.

Buyers span banks, lenders, payment processors, fintechs, card issuers and insurers, with the same underlying data serving risk teams and commercial teams through different workflows. Attribute coverage is unusually rich for business data, extending past identity and firmographics into monthly and annual card revenue, growth rate, average transaction size, payment technologies deployed and sub industry classification. The limit is geographic and currently absolute: this is United States coverage, with European and Latin American expansion stated as a 2026 objective rather than a present capability.