Financial Analytics · Finance Data Warehouse

Your reports disagree because each one reads its own copy of the ledger. We build the one layer they all read from.

A governed finance data warehouse for multi-entity groups, on any ERP, planning system and data platform: one entity key across every source, four layers from raw loads to certified finance models, and controls that check the numbers every day. Built to SOX-grade audit standards from day one.

90 daysTo first release

6 monthsTo full production

500+Tables integrated from six systems

4Governed layers, raw to presentation

At a global professional services group with 324 legal entities.

Book a 30-minute architecture walkthrough See the controls

The situation

Does this sound familiar?

Every team pulls its own extract from the ERP, so the same number has three values depending on who you ask.
Reports were built straight on raw tables, and nobody can say which transformation produced a figure.
An acquisition means months of mapping before the new business shows up in the group numbers.
The group already runs a data platform, and finance still does not sign from it. Every team has its own Excel jockeys, rebuilding the same numbers their colleagues in the next entity are rebuilding at the same time.

What is inside

Eleven report views on one governed layer

The views the platform ships with, grouped by the five areas of the finance function. Every view reads only from certified finance models. See all five areas

Reconciliation & Close

Statements & Reconciliation. The income statement and balance sheet on each ledger layer, with the reconciliation control.
Consolidated Revenue Bridge. ERP revenue to consolidated revenue, leg by leg.

See Reconciliation & Close

P&L & Management Reporting

Group Performance. Growth and profitability by entity, and how many group entities serve each client.
Department & Office P&L. Margin by department and office, with the loss-makers in plain view.
Job & Client Profitability. Best and worst jobs, billing realisation and client concentration, for project-based firms.
Entity ROIC. Where the return on capital comes from, entity by entity.

Balance Sheet & Working Capital

Work in Progress. WIP that ties to the general ledger, and work done but never billed.
Receivables. Ageing, with trade separated from intercompany.
Payables. Ageing by vendor, and a sub-ledger that ties to its control account.

Forecasting

Budget, Forecast & Outlook. Budget, forecast and actual side by side, variance by driver and the full-year outlook.

Controls and the AI layer

Controls & Data Quality. Six controls computed against the warehouse.
AI layer. Natural-language questions and variance commentary on the certified numbers.

For manufacturers, the same layer carries SG&A with allocations, product and customer margin, order to cash and inventory.

Works with

Built on the systems you already run

Any ERP, any planning or consolidation system and the operational systems around them, on any data platform. Nothing needs replacing first.

ERPs

SAP S/4HANA and ECC, Oracle Fusion Cloud ERP and E-Business Suite, Microsoft Dynamics 365, NetSuite, Infor, Sage and Maconomy, plus entities that still report on files.

Planning and consolidation

OneStream, Oracle Hyperion and EPM Cloud, Anaplan, SAP BPC, Workday Adaptive Planning, CCH Tagetik, Board and Jedox.

Operational systems

CRM such as Salesforce, HubSpot and Dynamics 365, warehouse management such as SAP EWM and Manhattan, project management such as Jira, Microsoft Project and Workfront, and HR and payroll such as Workday and ADP.

Data platform

On-premise SQL Server or Oracle, Databricks, Snowflake, BigQuery, Microsoft Fabric, Azure Synapse and Amazon Redshift.

Two ways to build it: we build the governed layer on the platform you choose, or we build the finance data products inside the platform you already run, in its catalogue and its deployment pipeline, with independent testing of the platform itself.

The architecture

One governed ledger between your systems

Every ERP and the group planning platform load into one governed ledger. A reconciliation control compares them on every load. Reports and the AI layer read only from the presentation layer.

Built with one entity key and one chart of accounts, column-level lineage generated by the build, infrastructure as code, one deployment pipeline through development, test and production, and federated identity with no stored secrets.

What it looks like

Six controls, computed against the warehouse

Illustrative statuses. Control names and definitions follow the Controls & Data Quality view.

Four ways consolidated numbers go wrong while looking right

Adding the ledger layers together counts every transaction twice. Layer integrity catches it.
A plain join on entities drops the elimination entities and overstates group revenue. Elimination integrity catches it.
Joining on the wrong account key moves amounts to the wrong lines. One conformed account key prevents it.
Leaving out the manual adjustments made in the group planning platform means the warehouse never matches the signed number. They load as their own ledger layer, reconciled and kept with their source.

Why build it

What it pays back

Results from our largest finance reference, a global professional services group with 324 legal entities.

4 daysFaster monthly close

50 to 12Hours for the monthly full load

4,000+Hours a year of variance commentary, now written by the AI layer

$57MCash released

The $57M was collected by an accounts receivable and work-in-progress programme that ran on the platform. Read the full case study

The first release is the start. The same layer carries a catalogue of 26 plays, from the shared identity layer to acquisition onboarding and tax pack automation. Each one is a small project with its own business case, typed by the value it brings: recurring cost reduction, cash and financing, revenue, risk avoided or acquisition readiness.

Global manufacturing group

After a major acquisition

Direct ERP ingestion replaced a manual extract relay.
More than 80% of data objects shared across reporting domains.

Global financial trading institution

Legal entity hierarchies, mastered

500,000+ master data records in 12 weeks.

Read the case study

Ask for a reference call under NDA

Questions

Questions a CIO asks.

Fit and scope

We already have a data platform. Can you build on it?

Yes. We build the finance data products on the platform you already run, inside its catalogue and its deployment pipeline, and we can test the platform itself independently.

Which systems and platforms do you work with?

The ones your group already runs. ERPs such as SAP, Oracle, Microsoft Dynamics 365 and NetSuite, planning systems such as OneStream, Oracle Hyperion and Anaplan, and data platforms on-premise or on Databricks, Snowflake, BigQuery and Microsoft Fabric. The layers, controls and deployment pipeline follow the same pattern on each.

Proof and delivery

How do you keep it SOX-compliant?

Infrastructure is defined in code, every change moves through one deployment pipeline, access is federated with no stored secrets, and every load leaves an audit trail. Our largest reference was built to SOX requirements from day one.

How long does it take?

At our largest reference, the first release went live in 90 days and full production followed in six months. The timeline scales with the entity count and the number of source systems.