From Synapse to Databricks: Data Warehouse Modernization at MOL Group’s Digital Factory
Client
MOL Group is one of the largest companies in Central Europe and the largest company in Hungary. Headquartered in Budapest, it is an integrated international oil and gas company with more than 80 years of industry experience, employing around 25,000 people and operating in over 30 countries.
Abylon has been a long-term partner of MOL Group, delivering multiple data and BI projects over the years. This project was carried out by MOL Group’s Digital Factory for MOL’s Retail division, which is responsible for operating 2,300 petrol stations and related convenience stores across 10 markets in Central Europe.
Challenge
MOL Group’s Digital Factory, the company’s digital transformation engine, previously relied on an on-premise, SQL-based data warehouse that primarily supported the business via BI reporting building on the vast amount of transactional and customer data. While stable and reliable, the system was not designed for advanced analytics or flexible data processing.
As demand for deeper insights, advanced solutions and modeling grew, an advanced analytics platform was introduced, combining technologies such as Databricks and Azure Synapse. While this enabled new capabilities, it also resulted in a fragmented environment, where different platforms were used by different user groups for development (Databricks) and analytics (Synapse) tasks.
Running multiple platforms increased complexity and cost while creating inconsistencies in performance and usage. The lack of a unified, cloud-based data platform limited scalability and made it harder to support advanced analytics and machine learning initiatives.
MOL Group’s Digital Factory needed a single, scalable platform for developers and analysts that would improve performance and reduce complexity.
Solution
As part of earlier initiatives, an advanced analytics platform had been established, combining Databricks for development and machine learning use cases with Azure Synapse for SQL-based analytics. While this setup enabled new capabilities, it also introduced a split environment where different platforms were used by different user groups.
As the platforms evolved, Databricks introduced SQL Warehouse, a dedicated SQL engine designed for analytical workloads. This made it possible to bring development and analytics onto a single platform without compromising performance or functionality.
An initial evaluation showed significantly better query performance and lower costs with Databricks SQL Warehouse. Based on these findings, a joint decision was made with MOL Group’s Digital Factory to migrate all SQL workloads from Synapse to Databricks, as it proved to be the most effective solution for both analytical and development use cases without requiring changes to existing code.
The migration covered queries, reports, and user workflows. After the transition, Synapse was decommissioned, and Databricks became the single platform for both developers and analysts.
A key part of the solution was the introduction of Databricks SQL Warehouse. This SQL-focused engine allows users to execute queries directly within Databricks and connect reporting tools such as Power BI through native connectors, enabling existing reports to be reused without significant rework.
Databricks is now the standard platform for analytics and development. All data processing and analytics now run in a single environment, simplifying operations and providing a scalable foundation for future use cases.
Results
The new platform provides MOL Group’s Digital Factory with a more efficient and scalable foundation for data-driven decision-making. By consolidating systems and improving performance, the organization reduced costs while enabling faster, more reliable access to insights. At the same time, the unified platform simplified operations and created a strong foundation for future data initiatives.
Key achievements:
- Reduced overall platform costs by approximately 60–70% compared to the previous setup
- Faster access to insights supported by improved query performance
- Unified development and analytics on a single platform, eliminating parallel environments
- Improved collaboration across teams through a shared data and analytics environment
- Reduced data silos and improved data consistency
- Simplified operations and governance with one central data platform
- Enabled advanced analytics, machine learning, and AI use cases on a scalable foundation
- Improved user experience for both analysts and developers
- Built-in version control and integration with Azure DevOps and Git, supporting more efficient development workflows
More Docuflow Case Studies
Project Type
Data Warehouse Modernization and Cloud Platform Consolidation
Industry
Oil & Gas Retail
Technologies
- Databricks
- Databricks SQL Warehouse
- Microsoft Azure
- Azure Synapse Analytics
- Power BI
- Python
- Machine Learning
