How a Global Manufacturer Built One Source of Truth for Product Data

A global manufacturer held its product master data in four separate enterprise systems. Marketing, sales and supply chain each had access to a version of the truth, which meant that a question about a product specification was answered differently depending on who you asked. OnPoint Insights consolidated the data into an Azure platform with a layered model, built a tool that lets any team query it directly, and controlled access by role.

4 systems, 1 source of truth

Product data from four enterprise platforms consolidated, so business teams query one place instead of deciding which system to believe.

One tool, any search criteria

A web-based tool lets marketing, sales and supply chain search product data on multiple criteria and gets results from across systems in one response.

Unified, not wide open

Active Directory integration means one shared source of product data still reaches only the people entitled to see it.

The problem

Four systems, four answers

Master data problems rarely announce themselves. They turn up as a marketing team quoting one specification, a supply chain team working from another, and a long meeting about which is correct. The cause is always the same: the product exists in several systems, and none of them is the authority.

Our client’s product data was distributed across its ERP, CRM and product information platforms, each maintained by different teams for different purposes. Business functions could not get consistent, trusted information, and the data was sensitive enough that simply opening everything up to everyone was not an option.

What was getting in the way

  • Product data was spread across multiple enterprise systems, creating silos and limiting visibility across business functions.
  • The transformation logic was complex, and it needed validation and quality checks at every stage rather than at the end.
  • Access had to be secure and role-based, so sensitive product data reached only authorized users.
  • Components and artefacts had to move reliably from development through testing to production, every time.

Our approach

Consolidate, layer, then let people in

What we found

Unifying the data is the straightforward half. The harder half is that a single source of truth is only useful if people trust it, and trust comes from validation they can see and from a structure where the approved version of a record is obvious rather than negotiated.

What we did

One place to ask the question. A web-based tool lets users query product data on multiple criteria and retrieve results from across all the source systems in a single response.

Consolidated, then layered. Data lands in Azure Data Lake and is structured in Azure Databricks through three layers, raw, cleaned and business-ready, so the version a team should be using is never in doubt.

Orchestration that handles both load types. Azure Data Factory runs the pipelines on schedule with dependency management, covering both full and incremental loads.

Access controlled by role. Active Directory integration enforces role-based access, so the unified view does not become an open one.

The choice we made

We built the deployment pipelines before anyone needed them. Automating the promotion of Data Factory components and database artefacts through development, test and production looks like overhead at the start of a project. It is what keeps a master data platform correct after the consultants leave: every change moves the same way instead of being reapplied by hand and drifting. Master data that has drifted is worse than no master data, because people have already believed it.

BUILT ON
  • Azure Data Factory Azure Data Factory
  • Azure Data Lake Analytics Azure Data Lake Analytics
  • Azure DevOps Azure DevOps
  • Azure Synapse Analytics Azure Synapse Analytics
  • Azure SQL Database Azure SQL Database
  • Microsoft SQL Server Microsoft SQL Server
  • Databricks Databricks

The results

What changed for the business

Product information stopped being a question of which system you trusted.

One answer across the business

Marketing, sales and supply chain work from the same product record, so a specification question has an answer rather than a meeting.

Accuracy that can be checked

Automated validation and QA, with a layered model behind it, mean a figure comes with provenance rather than with an assurance.

Unified without being exposed

Role-based access means consolidating the data did not mean broadcasting it, which is the objection that stops most master data projects before they start.

Let's Connect

Facing a similar challenge?

Talk to our data and AI team about where you are today and what good looks like. 30 minutes, no obligation.

Book a consultation
Professional customer service agent on a support call
More Case Studies
Next Case Study
Contact Us

Collaborate with us

We're here to answer your questions and help you find the right solution.

Client-oriented
Results-driven
Problem-solving
Transparent

"*" indicates required fields

This field is for validation purposes and should be left unchanged.