Landsbankinn needed to streamline its data infrastructure and integrate more than 45 data sources, so implemented a logical data warehouse using the Denodo Platform. Davíð Jóhannsson, Head of Information Intelligence at Landsbankinn, said that at the heart of the project was the company’s desire to create a data sharing culture throughout the organisation. The Denodo solution helped it to achieve a consistent view across all data silos and today the entire bank is processing data from one place.
Landsbankinn hf. is the largest financial institution in Iceland. It was founded in 2008 and is the successor to Landsbanki Íslands, established in 1886. Its largest shareholder is the Icelandic State Treasury (98,2%). It boasts 40% of the retail and 32% of the corporate banking market share in Iceland.
Landsbankinn wanted to accelerate its key operational and BI reporting capabilities and provide real-time data for informed decision-making across all kinds of users in the organisation. To streamline its data infrastructure and integrate more than 45 data sources, Landsbankinn implemented a logical data warehouse using the Denodo Platform.
Business need
Before implementing a logical data warehouse, Landsbankinn had a large number of data sources and three main reporting platforms for reporting including general reporting, risk reporting and KPI reporting to top management and board. The total number of data sources at Landsbankinn is around 45, including five Oracle traditional databases, two data warehouses and around 10 Microsoft SQL Server databases. On top of this, flat files, Excel files, XML files, APIs from internal and external data sources are also used.
Despite having traditional BI and a data warehouse in place, users were often not able to find and access the data they needed. Business rules and logics were scattered, creating trouble for obtaining the lineage of the data. The large, extensive query points and business objects made the situation worse. Also, access controls restricted access to a database, or certain columns of the database, making it arduous to identify the problem.
These experiences led Landsbankinn to the idea of housing all business rules and security protocols together. But to consolidate these data sources and feed the demand of the various user groups such as the board of management, customers, general reporting, risk reporting and operational reporting, Landsbankinn needed a novel data integration solution that was cheaper, easier to implement and technology agnostic.
Solution
The large number of disparate data sources at Landsbankinn needed to be integrated to build a data-driven, holistic overview of business activities, so the company could use that integrated data to optimize operations and improve the customer experience. Landsbankinn implemented a logical data warehouse (LDW) using the data virtualisation capabilities of the Denodo Platform. The LDW logically aggregates data from the disparate data sources, transforms the data based on the applied business rules and makes it available to the consuming application through a variety of APIs, such as JDBC, ODBC and REST. The LDW at Landsbankinn was implemented in three phases:
1. In the first phase, data virtualisation was used to create a single data access layer to the unified enterprise data, sourced from a variety of disparate source systems, and make it available to reporting tools, analytics and APIs.
2. In the second phase, data virtualisation capabilities were expanded by connecting new BI and analytics platforms to the data virtualisation layer, enabling business systems to query the data virtualisation layer, and adding more data sources including flat files, SaaS platforms, National Statistical data, Central Bank data and some open data platforms as well.
3. The third phase involved creating a data sharing platform (DSP) in which a domain developer working on the source database itself would create the sharing view. These views can be accessed by users working in different domains. This data sharing platform makes it easier for the information intelligence team to know what a database contains.
Landsbankinn is leveraging the LDW to streamline its data infrastructure in a variety of ways:
• The overarching, central LDW is now used primarily for traditional BI and all the data preparation now takes place in the LDW.
• A self-service analytics platform has been built, using Tableau, which heavily uses the Denodo Platform’s data catalogue to identify the relevant datasets among the shared datasets.
• Web services such as REST APIs are being used to connect natively to Microsoft Excel to retrieve data and enable users to experiment. Excel is still heavily used for ad-hoc data extraction at Landsbankinn.
• APIs are also being used for connecting operational systems, such as CRM systems, to the warehouse.
Benefits
- The logical data warehouse has extended the reach of data across the organisation. Around 80% of the targeted users are now using the LDW in one way or another. Adoption for the remaining 20% is in full swing.
- The data can now be used by all kinds of users within the organisation, even the ones with limited data/IT skills, to make informed decisions.
- Landsbankinn gained significant improvement in terms of time and resources required to perform operational reporting and traditional BI reporting.
- Data security is now revamped in a central, global repository that is the LDW, using row-level restrictions and corporate directories.
- Self-service analytics was extensively used to provide specific dashboards relative to the COVID-19 crisis, for the board and managing directors.
Davíð Jóhannsson, Head of Information Intelligence at Landsbankinn, expands on how it collaborated with Denodo to achieve its business goals and how it has been able to build a data-driven, holistic overview of business activities.
Can you tell us more about how real-time data is used for informed decision-making across the organisation?
It is extremely important that there is good access to new and correct data when making decisions across Landsbankinn – whether they are strategic or financial. The bank is increasingly providing self-service loans and it must be possible to integrate and make decisions based on data that is updated in different places. We also use real-time data in marketing campaigns, enabling us to quickly change tactics when needed. When data can be viewed in real time, new opportunities arise to enhance data-driven decision-making.
Why did you decide to work with Denodo on this project and did you achieve what you set out to?
At the heart of this project was our desire to create a data sharing culture throughout the organisation. We wanted to accelerate our key operational and reporting capabilities through providing real-time data to all users to help inform decision-making and improve both operational efficiency and customer experience initiatives.
Before working with Denodo to implement a logical data warehouse, we had over 45 siloed data sources – including Oracle databases, data warehouses and APIs from internal and external sources. This was making it increasingly complex for our users to find and access the right data at the right time.
We’d already made some changes – housing all business rules and security protocols together – but to consolidate these data sources and feed the demand of various user groups we needed a novel data integration solution that was more cost-effective, easier to implement and technology-agnostic. This is where Denodo came in.
The Denodo solution helped us achieve a consistent view across all data silos and today the entire bank is processing data from one place.
How did the implementation of a logical data warehouse using the Denodo Platform help you to streamline your data infrastructure and what benefits did this bring?
Our team used the Denodo Platform – a data integration and data management solution built on data virtualisation – to build a logical data warehouse. Through this self-service platform, we are able to aggregate data from disparate data sources, transform that data based on the applied business rules and then make it available to consuming applications. Essentially, this means that throughout the bank, data can be accessed and analysed by all kinds of users, even those with limited IT skills. It enables our business leaders to make informed decisions and deliver a vast range of services to Landsbankinn customers quickly and efficiently.
What prompted Landsbankinn to expand its approach from building just a logical data warehouse to building a data mesh architecture and what would you recommend to other organisations building a data mesh architecture or attempting to?
The bank wanted to become data-driven, which means that the entire bank needs to be data-driven, not just individual departments. The traditional ETL route is not suitable for this kind of strategy as it scales very poorly. Instead of increasing the staff in the warehouse team, it is better to make better use of programmers, transferring more responsibility to them and thereby increasing data to market. Data mesh is ideally suited for crowdsourcing access to data. The management of the data is then in the right place, which makes the data more reliable. We also get rid of unnecessary processing time when the data is being moved to a warehouse as well as duplicating the data, but this entails the risk of errors. When companies are considering data mesh, it needs to be part of a long-term strategy and everyone needs to buy into this architecture. Whether it’s analysts, programmers or DBAs, data mesh won’t work unless there’s good cooperation between everyone and everyone is working towards the same goal.
Can you tell us about some of the overall benefits achieved since working with Denodo and what this means for your Digital Transformation journey going forward?
The logical data warehouse has extended the reach of data across the organisation. In fact, around 80% of the targeted users are now using it in one way or another and adoption for the remaining 20% is in full swing. This has resulted in some huge benefits, including:
• Our data can now be used by all kinds of users within the organisation, even the ones with limited data/IT skills, to make informed decisions.
• Significant improvement in terms of time and resources required to perform operational reporting and traditional BI reporting.
• Our data security is now revamped in a central, global repository that is the logical data warehouse, using row-level restrictions and corporate directories.
• Self-service analytics was extensively used to provide specific dashboards relative to the COVID-19 crisis, for the board and managing directors.
The Digital Transformation has happened more quickly as Denodo has provided rapid feedback on the digital solutions as soon as they are released and we can then see how we are getting customers to use the solutions. Today, it is standard in all new solutions that there is real-time access to the underlying data.
