The Official Source of Truth for Corporate Banking

Rely on official government registry data to access up to 20 years of financials, map complex parent-subsidiary networks, and resolve entities globally.

Searching registries01 / 04
Resolving ACME Logistics Holding B.V.
🇬🇧Companies House · United Kingdomchecked
🇩🇪Handelsregister · Germanychecked
🇫🇷INPI · RNE · Francechecked
🇳🇱KvK Handelsregister · Netherlandschecked
🇪🇸Registro Mercantil · Spainchecked
🇮🇹Registro Imprese · Italychecked
🇵🇱Krajowy Rejestr Sądowy · Polandchecked
🇷🇴ONRC · Romaniachecked
🇬🇧Companies House · United Kingdomchecked
🇩🇪Handelsregister · Germanychecked
🇫🇷INPI · RNE · Francechecked
🇳🇱KvK Handelsregister · Netherlandschecked
🇪🇸Registro Mercantil · Spainchecked
🇮🇹Registro Imprese · Italychecked
🇵🇱Krajowy Rejestr Sądowy · Polandchecked
🇷🇴ONRC · Romaniachecked
0 registries · 0 countries · 0M companies indexed
Identity confirmed at source
Legal nameACME Logistics Holding B.V.filed
Registration🇳🇱 KvK 33027-BNL
StatusActivefiled
Incorporated22 March 2016filed
Turnover FY25€ 214,780,000filed accounts
Source KvK Handelsregister · retrieved 04:12 UTC
Ownership resolved to natural persons
ACME Logistics Holding B.V.
Meerkamp Beheer B.V.58.0%
J. Meerkamp54.2%
Effective beneficial holding0.0%
2 links evidenced · no gaps · UBO-register and shareholder filings
Screening entity, officers and beneficial owners
OFAC SDN and Consolidated USno match
EU Consolidated Financial Sanctions EUno match
UK OFSI Consolidated List UKno match
UN Security Council Consolidated UNno match
Clear · 50% ownership aggregation applied
Run 29 Jul 06:02 UTC · list versions recorded for replay
synced dailyprovenance returned on every field
0+

years sourcing registry data

Sourced directly from government registries

Source name and link on every dataset

Retrieval timestamp on every field

The original filing attached

Trusted by compliance, risk and data teams at

allianz logo
uber logo
amazon logo
statista logo
lseg logo
persona logo
zeizz logo
sap logo
schneider logo
kroll logo
experian logo
quantifind logo
bdo logo
allianz logo
uber logo
amazon logo
statista logo
lseg logo
persona logo
zeizz logo
sap logo
schneider logo
kroll logo
experian logo
quantifind logo
bdo logo
allianz logo
uber logo
amazon logo
statista logo
lseg logo
persona logo
zeizz logo
sap logo
schneider logo
kroll logo
experian logo
quantifind logo
bdo logo
allianz logo
uber logo
amazon logo
statista logo
lseg logo
persona logo
zeizz logo
sap logo
schneider logo
kroll logo
experian logo
quantifind logo
bdo logo

Three teams. Three ways a company record gives way.

Different systems, one underlying question. Who is this entity, who stands behind it, and can we show our working twelve months from now.

01

Corporate onboarding and CDD

A UK entity gets onboarded on a name match against a register entry nobody behind it ever had to prove. Under the Economic Crime and Corporate Transparency Act those directors and PSCs are being verified one by one, and the transition closes on 18 November 2026. Records that looked settled at onboarding are moving underneath the file right now.

What closes it

The PSC statement as filed, officer appointment and resignation dates, and a daily check against the register that tells you the day any of it changes.

02

Credit origination and lending

The borrower hands over a spreadsheet of group entities. The registries say there are more. Exposure aggregated on a structure the counterparty declared about itself is exposure the bank has not actually measured, and the gap only shows up when something defaults.

What closes it

Subsidiaries and branches resolved from filings across every jurisdiction the group touches, with the ownership percentage evidenced on each link.

03

Correspondent banking and trade finance

The intermediary two layers down sits in a jurisdiction where the register is not in English and not online. That is precisely where the ownership or control test fails, and it is the layer most datasets quietly stop at.

What closes it

Direct sourcing in the jurisdictions that are hard to reach, with aggregation run through the whole chain rather than the first layer of it.

Six data domains, one provenance model.

Sourced from the registries that issued the entities, not assembled from resellers. Every field in every domain carries the same evidence trail.

Registry data

Legal name, registration number, legal form, status, incorporation date and registered office as the issuing registry holds them. Previous names, officers with appointment and resignation dates, activity codes, and the full filing history since incorporation.

Financials

Turnover, operating profit, total assets, equity, cash and employees taken from the accounts the company actually filed, with the reporting period, currency and accounting standard attached. Derived figures such as net debt are labelled modelled.

Ownership and UBO

Direct shareholders with percentages and share classes, effective ownership calculated through every intermediate holding, beneficial owners above threshold, PSC nature-of-control statements and voting rights. Where a register restricts access, the gap is disclosed rather than inferred.

Group structure

Global ultimate parent, domestic ultimate, every subsidiary and registered branch, resolved across jurisdictions so credit exposure and screening aggregate at group level. Each node carries its registry number, its LEI where issued and a stable identifier.

PEPs and sanctions

Screening applied to the entity, its officers and its beneficial owners, with aggregation run across the ownership chain. Every match returns the list entry, the regulation reference and the publication date. Every run records the list versions used, so it can be replayed.

Sources, links and timestamps

The source registry, the official filing it was read from, a link that opens the document, the filing date and the retrieval timestamp to the second. Filed and modelled are labelled separately. Prior values are retained with the date they were superseded.

The score, and everything under it.

Scores and limits are modelled. Judgments and filed accounts are not. Both arrive labelled, so a credit committee can see which part is our opinion and which part is on the public record.

01
Credit scoreModelled from filed accounts, payment behaviour, judgments and group structure, with the drivers exposed rather than sealed inside a black box.
02
Credit limitA recommended maximum exposure with the inputs shown, so your own model can override it rather than inherit it.
03
Trade payment behaviourDays beyond terms and payment trend, drawn from contributed ledger data and labelled as such.
04
County Court JudgmentsCase number, amount, date and satisfaction status from the official register of judgments for England and Wales.
KESTREL MACHINING LTDGB · 07412886 · sample record
Assessing
Credit score
Low riskMODELLED
filed accountspaymentsjudgmentsgroup
Recommended credit limit
£ 0MODELLED
Trade payment behaviour
0 daysbeyond terms · improvingCONTRIBUTED
FebMarAprMayJunJul
County Court Judgments
1£ 18,400 · unsatisfied · Mar 2025FILED
Filed sits on a document. Modelled shows its method.

Perpetual monitoring across the whole counterparty book.

A counterparty cleared at onboarding can be sanctioned by lunchtime. A periodic review would find it at the next cycle. Every entity here stays under watch from the day the relationship opens, re-checked against its registry daily, with changes surfaced on the day they are filed.

Counterparty book watch2,418 entities · 41 jurisdictions · continuous
00:00:00 UTC
Sample streamSelect a row to see the filing behind it
Case study · Experian

Experian Enhances Benelux Company Data Coverage

Experian expanded verified company data coverage across the Netherlands and Belgium, strengthening the credit-risk and fraud-prevention signals it delivers to its own banking clients.

Read the case study
How we source

From the filing to the field.

Anyone can claim a registry as a source. The work sits in what happens between the filing being published and the value arriving in your system.

StageWhat happens
01Global standardisationData from 400+ registries is harmonised into a single schema. Field names, date formats, currency codes and legal form types are normalised across every jurisdiction.Besloten vennootschaplegal_form: BV · NL
02Language translationCompany names, addresses and officer details are translated from the local language into English, with the original filing language preserved alongside the translation.Handelsnaam · BestuurderTrading name · Director
03Financial digitisationScanned PDF filings are read into structured balance sheets, profit and loss statements and ratios, so accounts filed as images become data.jaarrekening_2025.pdfturnover · assets · equity
04Firmographic enrichmentMODELLEDRevenue estimates, headcount, industry tags, SIC and NAICS codes, technology stack and digital footprint are layered on top of the statutory record.registry record+ employees · industry · web
05Corporate linkage and entity resolutionParent and subsidiary trees mapped across 40M+ companies. Cross-border resolution connects the same company or the same person across separate jurisdictions.3 records, 3 countries1 group, ultimate parent resolved
06Integrity and deduplicationDuplicates removed, addresses geocoded, industry classified and company-to-website links verified before anything is published.4 near-duplicate records1 entity, geocoded and classified

How you get the data.

Sourced from 400+ government registries across 200+ countries. Every record attributed to the filing it came from.

Platform

Search, reports and monitoring for analysts working cases directly.

Regis

AI. Ask the whole dataset in plain language. Regis answers on registry data and returns the filing behind every number, so the answer can be checked rather than trusted.

API

MCP server. Instant KYB, financials, ownership and credit endpoints, plus registry data inside any AI agent.

Data feeds

Bulk datasets, enrichment and redistribution licensing to power your own product.

CRM integrations

Enriched records and monitoring inside the systems your teams already work in.

AI connectors

Registry data inside your AI assistant, under your own access controls.

Test it on a hundred counterparties you have already onboarded.

We return them with the registry, the filing, the document link and the retrieval timestamp on every field, and flag anything that has changed since the day you onboarded them.

Before procurement, the same ten questions come up.

Where does the data actually come from?

Directly from government registries and official filing sources, over 400 of them across more than 200 countries. Every field carries the registry it came from. We provide the source list per country before a contract, not after.

How current is a given record?

Registry records are synced daily and each field carries the exact time it was retrieved. Freshness is a property of the record rather than an average across the dataset, so you can see which specific fields are stale instead of trusting a headline cadence.

Can we see the underlying filing?

Yes. Document references resolve to the filing the field was read from, so an analyst or an auditor can open the source without leaving the case.

How do you handle beneficial ownership where the register is closed?

We evidence what the jurisdiction permits and state the gap explicitly where it does not. Percentages are calculated through the chain from filed links, and every link carries its own evidence. We do not fill gaps with inference and present the result as fact. A disclosed gap is something a compliance officer can work with.

Which lists do you screen against, and how is ownership aggregated?

OFAC SDN and Consolidated, the EU Consolidated Financial Sanctions List, the UK OFSI Consolidated List and the UN Security Council Consolidated List, among others. Screening runs against the entity, its officers and its beneficial owners. Aggregation is applied across the ownership chain, including the OFAC 50 percent rule and the EU and UK ownership or control tests.

Can we reconstruct a decision made two years ago?

Prior field values are retained with the dates they were superseded, and screening runs can be reproduced against the list state in force at the time. That is what makes a lookback defensible rather than reconstructed from memory. Retention windows are agreed at contract.

How do you match this against the entities already in our book?

Records carry the registry number and the LEI where one has been issued, alongside our own identifier. You can join on whichever key your systems already use. For an existing portfolio we run a matching pass over your file and return the resolved entities with the evidence for each match, so your team reviews the joins rather than accepting them blind.

How do we hear when something changes?

Change events are pushed the morning after they appear on the register. Status changes, officer and ownership movement, new accounts, share capital changes, charges and fresh list matches all carry an event type, so they can be routed to case management, credit review or financial crime alerting separately rather than landing in one queue.

How do we integrate, and what does implementation involve?

Start on the API against live endpoints with a trial key and no implementation work. Production delivery is by REST API, signed webhooks, bulk and delta files over SFTP or object storage, or connectors for Salesforce, HubSpot and Dynamics 365. Bulk and residency arrangements are scoped with your data and security teams.

What are the licensing terms for group-wide use?

Licensing is agreed at institution level, including internal reuse across entities and onward distribution where relevant. We would rather agree scope once than meter every look. Registry numbers and LEI travel with the record, so nothing forces you to stay.