Smarter and faster decision-making with all data in one platform
“By providing secure, scalable and reliable infrastructure, thus facilitating advanced analytics, Databricks plays a key role in fulfilling our vision of enabling stakeholders to make more efficient and informed decisions.”
Swedbank is the largest bank in Sweden and the third-largest in the Nordic countries, with more than 7 million customers and approximately 15,000 employees. Five years ago, the bank started looking at how to integrate and analyze data and use insights to improve decision-making. To assure a stronger position in the market, with reduced costs, increased revenue and enhanced customer experiences, Swedbank migrated existing capabilities and application services to Databricks’ driven analytics platform within the Azure cloud and implemented Immuta as its unified access control plane.
Unique challenges in the banking industry
Compliance, legal requirements and security are the main challenges banks like Swedbank face in terms of handling data.
“It’s not easy,” says Vineeth Menon, Head of Data Lake Engineering at Swedbank. “The biggest challenge, in my opinion, is overall trust in the data collected and the results achieved through analysis of this data. Problems with data quality also remain at the forefront. And there are several other challenges, such as data silos and a lack of organizational maturity to be data driven.”
The key to the whole project was a risk-based approach, which involved analyzing and evaluating each stage of the migration with a feasibility perspective.
“The migration strategy was designed with a twofold sourcing approach, meaning we could still have on-premise scenarios continuing in parallel with the data migration to the cloud platform,” says Menon.
Vision: To bring value across the organization
To really get the best out of a cloud-based platform set for analytics, it’s important to make a cultural shift within the organization. Naturally, as part of a legacy industry with close to 200 years of history, this was new territory for Swedbank.
“The vision was to bring real value into the project. We knew it wasn’t possible for everyone to be involved in everything at once. Our strategy was to always prioritize bringing maximum value by making each individual’s work more time-efficient. Thanks to Databricks, we can now provide easy access to onboarding, support, training, and a shorter time-to-market for each use case. We unlock potential value across the entire organization by creating a sharing culture,” says Menon.
“The interesting thing about this project is that Swedbank can now manage markets and share trading, customer analysis, and cybersecurity in one single internal data platform. Since the framework has been developed, they can create customized solutions to meet all requirements in-house. Databricks is so flexible that it provides a wealth of possibilities,” explains Simon Ghebremikael, Strategic Enterprise Account Executive at Databricks.
AI and ML available in one click
Swedbank’s prior on-premises solution worked and served its purpose for the most part, but it had major downsides. Mandatory investment for on-premises hardware, major upgrades, and the effort involved in migrating existing services were time-consuming and led to increased operational costs. It also had limited possibilities for separating storage from compute, and no flexibility to add third-party solutions for deep learning and AI.
“Our advanced analytics platform driven by Databricks has made artificial intelligence and machine learning programming languages available in one click through the centralized hub,” says Menon. “It has really helped us to increase our productivity. Implementation of machine learning ops using the platform has enabled shorter development cycles, which has resulted in shorter times-to-market. It also has the ability to detect and act on new patterns related to fraud, suspicious behavior, and so on, as well as providing us with the ability to utilize green cloud providers to minimize environmental impact.”
Implementing Immuta in its analytics platform has gained Swedbank complete control of who gets access to what data, and it provides a centralized mechanism for managing data governance without needing to duplicate it. Microsoft Azure has also played a significant role, in terms of having centralized production data availability for all the environments, for all of Swedbank’s customers.
Moving forward: Utilizing Databricks Data Intelligence Platform to its best potential
Swedbank’s vision for the future is to satisfy the need for advanced analytics in the bank, allowing stakeholders to make more efficient and informed decisions.
“Databricks plays a key role in fulfilling that need by providing reliable and powerful infrastructure, which simplifies parallelization and distributed computing. With decoupling of compute and storage, the need to move the data across different workspaces has reduced, resulting in faster access to data with lower costs involved,” says Menon.
Swedbank’s goal is to derive more value from the data so enterprise analytics adoption isn’t limited to just one or two business users, but is expanded across the bank. This will allow the analytics platform to be utilized to its best potential, serving the various businesses, their insights and goals.
“I believe everyone involved in this project is proud of how well and efficiently it was executed. It was a critical process, as the platform is so important, and it was also one of the fastest migrations Swedbank has ever done,” says Ghebremikael.
