Skip to content

Customer story

It started as verification.
It became fraud intelligence.

Ramp issues corporate cards and manages spend for the businesses it onboards, under compliance requirements set by its banking partners. Ramp sent 4,000 real applications to Enigma as a test: 84% matched on Enigma alone, 94% combined with Ramp's existing sources. What began as a verification check now runs as six products in production.

Enigma + Ramp — from business verification to fraud intelligence

Customer
Ramp
Category
Corporate cards & spend management
Where Enigma runs
Onboarding KYB, EDD, address intelligence
Headline result
6 products in production
A row of small commercial storefronts at dusk — a tyre shop, a bakery, a hardware store, a laundromat — interiors lit, people working inside, wet asphalt in the foreground
Every one of these is an applicant somewhere. The question is only whether the record says so.synthetic render

84% → 94%

match rate on a 4,000-application test — Enigma standalone, then combined with Ramp's existing sources

6

Enigma products running in production at Ramp

~10x

growth in Ramp's application volume over the engagement

~87%

of Ramp's manual reviews triggered by address failures

Source: Enigma + Ramp case study, enigma.com — 4,000-application match-rate test

01

The challenge

A pipeline that works at one scale doesn't hold at ten times the volume.

Ramp's banking partners set the requirements and they are non-negotiable: validate the business, confirm the TIN, flag anything that poses fraud risk. When Ramp first engaged Enigma it was processing a few hundred applications a month. Today it is thousands, with plans to double again.

The cost of that growth lands in manual review. Every application that stalls there — not because the data doesn't exist, but because the pipeline can't find it — adds friction that compounds with volume.

Ramp's team saw it early. They wanted a provider with strong entity linking across verified sources — government filings, tax records, business registrations, key business directories — that could deliver better coverage, faster responses, and richer fraud signals.

02

The evaluation

Four thousand real applications, and two capabilities built from the results.

Ramp opened the conversation in November 2023 and sent 4,000 real applications for testing. Most KYB providers rely on Secretary of State data alone; Enigma's entity resolution links business records across SOS filings, government registrations, tax records and key business directories. The differentiator isn't the data on its own — it's the linking that makes it actionable.

The technical discussions went deep: Levenshtein distance thresholds for fuzzy name matching, address normalization, entity resolution across inconsistent state filing formats. Enigma then built two new capabilities directly out of the evaluation — address verification tasks and name verification tasks, calibrated to Ramp's input patterns.

Integration meant configuring a new data source rather than building a new one, since Ramp already routed verification through an orchestration partner. Enigma entered on a pay-for-results model — charges apply when a verified result comes back, not per API call regardless of outcome. Champion/challenger testing on live traffic validated Enigma against the incumbents before full cutover, and TIN/EIN verification was added during integration using real-time IRS matching through the same workflow.

Enigma standalone

84%

Combined with Ramp's existing sources

94%

  1. 01

    Discovery

    First call to evaluation kickoff

  2. 02

    Evaluation

    4,000-app test: 84% standalone, 94% combined

  3. 03

    Integration

    Champion/challenger testing on live traffic

  4. 04

    Production

    6 products live and counting

03

The result

The address was real. The business behind it wasn't.

Registered agent addresses are the single biggest driver of manual reviews in Ramp's onboarding pipeline. Businesses registered at CMRAs, registered agents and coworking spaces pass basic verification because the address is real — it just doesn't tell you whether there is a physical business behind it.

Enigma built virtual address classification for the integration: registered agent, CMRA and non-physical address detection. For a business with only a virtual registered address, Enigma can surface a verified physical address, turning a potential false positive into a resolvable case.

For enhanced due diligence the problem is different — each business is different, and investigations depend on specific risk signals. A compliance analyst on Ramp's KYC/BSA/CIP team built his workflow around Enigma's AI-enhanced research tool, a natural language interface over open-source intelligence and Enigma's Gov Archive: billions of federal, state and municipal records, including OSHA violations, business inspections, professional licenses and regulatory filings sourced directly from government agencies. Several members of the compliance team now use it, scaling capacity without proportionally growing headcount.

  • ~87%

    of manual reviews triggered by address failures

  • ~50%

    false positives in a sample of ~50 flagged addresses (Nov 2025)

  • 3

    further products in development or POC

We're enhancing our EDD program and looking at Enigma to help streamline it. The tool recommended not to do business with this customer, we can't even verify they exist, the beneficial owners don't exist. We make our own decision on the Ramp side, but it gave me the leverage and the sourced evidence to help in saying no.

Joshua Mungillo

Compliance, Ramp — October 2025

We know the process wouldn't scale efficiently at our growth targets. We're targeting a 50 percent reduction in manual review across the whole pipeline.

Alex De Jesus

Head of Fraud Management, Ramp — September 2025

Also in production

Run the same test on your book.

Bring the businesses you clear today and see what verifying against independent evidence changes.