Enabling real-time insights for smarter merchant decisions
Worldline unifies merchant and transaction data for streamlined reporting
Reduction in monthly infrastructure costs
Increase in schema reporting
Lift in team productivity
Worldline, the leading payment processor in Europe and ranked fourth globally, supports millions of merchants and banks across 50 countries. Offering multiple services from payment processing to currency conversion and financial analytics, Worldline wanted to pool more of their resources to focus on business intelligence (BI) for internal and external reporting and machine learning operations that would end “data wrangling” once and for all. The fintech company also wanted to start using transaction data and customer interaction analysis for churn prediction and anomaly detection. However, Worldline’s global expansion through acquisitions created a fragmented data landscape that made it challenging to unify data and scale analytics across 50 billion transactions and trillions of Euros. Legacy on-premises systems further hindered the company. To overcome these hurdles, Worldline partnered with Databricks, reducing infrastructure costs by €200,000 per month.
Dealing with disconnected systems from multiple acquisitions
Worldline’s ambition to provide comprehensive reporting data and insights to internal teams was critical for maintaining their position as Europe’s leading payment processor. For this reason, a major initiative, known as “Drive,” was designed to consolidate data from various sources and create a unified view of transactions, customers and merchants. This project aimed to equip Worldline’s management and business analysts with comprehensive dashboards that would monitor important key performance indicators (KPIs) related to financial reporting, forecasting and sales performance. The unified view would empower leadership and operational teams to make swift, data-driven decisions, while BI cubes — multidimensional data structures — could enable analysts to explore the data in depth, perform complex queries and create self-service dashboards. Teams could now forgo the usual round of IT intervention to make data analysis more accessible and cross-functional teams more capable in their respective roles.
When it came to their customer-facing solutions, Worldline also wanted to prioritize providing merchants with data insights that helped them benchmark performance, process refunds and visualize transaction trends. These services supported a range of clients — from small shops to major retailers and even airlines — and helped all of them use daily transaction data for better operational planning and internal reporting. Simultaneously, Worldline aimed to strengthen their data science capabilities to implement more sophisticated fraud detection and anomaly identification systems to protect these very clients. Such improvements would help the business identify and mitigate suspicious activity, protecting both merchant and customer data.
As the company grew, it was increasingly challenging to unify customer, transaction and merchant data — over 50 billion transactions annually. To put this into perspective, that equates to 10,000 transactions per second. As Stephan Pirson, Chief Data Officer, Merchant Services Division at Worldline, put it, “We needed to move from a patchwork of systems to something that could process data at scale and give us the ability to go from big-picture trends down to the smallest transaction details.” Ultimately, this is why the company decided to invest in the Databricks Data Intelligence Platform.
Creating a layered system for real-time business data
Worldline’s transition to a cohesive, scalable data platform addressed the unique challenges posed by their rapid expansion through mergers and acquisitions, which created a fragmented data ecosystem with disparate sources. At the core of their tech stack, Apache Spark™ and Databricks clusters were deployed to centralize and streamline data processing, managing the massive scale of Worldline’s operations, from data ingestion to real-time processing. Delta Lake’s ACID-compliant architecture ensured data integrity and reliability for both internal and external reporting, vital for handling over 50 billion transactions annually. “Databricks fundamentally transformed how we approach data,” explained Pirson. “With Databricks, we’ve centralized our data in one place, enabling us to scale analysis across millions of transactions while maintaining accuracy and speed. This shift lets us prioritize innovation, which helps us maintain our competitive edge in the market.”
Leveraging Delta Lake’s foundation, Worldline implemented the medallion architecture, a structured data framework that enabled compliance, accuracy and accessibility across their data lifecycle. In the Bronze layer, Worldline stored raw, unprocessed data from various sources, creating a traceable repository for reprocessing and regulatory needs. The Silver layer transformed and refined this data, ensuring it was optimized for operational reporting and analysis. The Gold layer aggregated and curated data, readying it for high-level business applications like internal KPI dashboards and customer-facing analytics. Layered atop Apache Spark and their Azure data lake, this architecture created an efficient pipeline that converted raw data into actionable insights for Worldline’s teams and clients.
With this foundation in place, Worldline integrated Databricks with Power BI through their Drive initiative, consolidating disparate data sources into unified dashboards for real-time KPI monitoring. To improve collaboration, Unity Catalog governed and documented data access, allowing over 85 data engineers and 200 analysts to explore and analyze data securely through shared Databricks Notebooks. Further enhancing their solution, Worldline adopted MLflow to manage machine learning model lifecycles, optimizing model training and deployment for AI-driven fraud detection and anomaly identification. Although their ML operations are still maturing, this setup allowed the data science team to shift focus from data wrangling to strategic modeling. Altogether, this cohesive Databricks suite enabled Worldline to evolve from fragmented workflows to a scalable platform ready to drive advanced analytics and AI initiatives.
Equipping teams to succeed while reducing infrastructure costs
Worldline’s adoption of Databricks has streamlined their data infrastructure and allowed their teams to pivot from infrastructure management to delivering actionable business insights. By centralizing data processing and governance with Databricks, Worldline’s data engineers and analysts gained more efficient access to data and reduced reliance on cumbersome on-premises solutions. This shift has also fostered better collaboration across teams, with analysts able to work directly with data in real time, perform in-depth analyses and even contribute to debugging processes alongside data engineers. The enhanced workflows have cultivated best practices across teams to accelerate project timelines and data-driven decision-making. “With Databricks, we’ve not only streamlined our data processes but also unlocked a new level of efficiency across teams. We’re now equipped to serve our merchants and internal stakeholders with more accurate, timely insights,” concluded Pirson.
From a quantitative perspective, the shift to Databricks has delivered major financial and operational improvements. Worldline is projecting a €200,000 monthly reduction in infrastructure costs, primarily from moving data previously stored on Cosmos DB to the more cost-efficient Delta Lake on Databricks. Operational efficiency gains are equally compelling, with productivity improvements of 40% and schema reporting speeds now 93% faster. This enhanced capacity to handle complex reporting and analytics at scale has positioned Worldline to maintain high service standards for both internal teams and external clients.
In addition to immediate impacts, Worldline’s long-term goals reflect a commitment to deepening data access and literacy company-wide. They plan to complete a comprehensive data integration by 2025 and roll out a data literacy program that encourages self-service analytics through Databricks. By fostering a data-literate workforce, Worldline aims to help their teams independently analyze and act on data insights. The company is also exploring more advanced AI tools, such as Databricks AI/BI Genie, to further democratize access to insights by empowering nontechnical users to perform natural language querying to drive business growth.
