Alior bank Data Expo 2024

Transcript

1. Alior Bank S.A. Dariusz Dudek Data Manager Alior Bank Mariusz Wądołowski Country Leader Teradata Poland Alior Bank S.A. | September 2024
2. 2 Business strategy 2023-2024 „Your bank of everyday life, your Bank for the future”: • Universal Bank focused on building a digital relationship with a client • Leader of Consumer Finance market in Poland​ • Sustainable growth of the business clients’ credit portfolio, particularly in SMEs segment and support for the energy transition In 2023, after a 2-year break, the Bank returned to the WIG-20 index In 2024 Alior Bank received approval to pay out a dividend November 2008 – launch as a new type of bank - combining the principles of traditional banking with innovative products A modern, agile organization, employing over 7.2 thousand people, over 520 branches (internal and external network combined) 4.4 million retail customers and over 200 thousand business clients. Total assets amount to 22.5 billion US dollars Guided by simplicity, teamwork and high ethics. Works with committed experts. As a local bank, it has a high level of autonomy and creates solutions based on employees' ideas Built on new technologies, leading to gradual expansion and improvement of the offer with new products and services Bank for everyday, Bank for the future
3. 2008 2009 - 2015 2017 2024 3 Evolution of Teradata DWH at Alior Bank Several Data Warehouse modernization projects Implementation of a backup environment (DR) of the Data Warehouse CRM implementation BI implementation Launch of the Bank Implementation of the production environment (PROD) of the Data Warehouse Obligatory Reporting Management Reporting Separation of analytical and reporting environments Tableau implementation Implementation of the "Business Self- Service" model Integration with the cold storage archival environment Combining of physical environments Migration to the Microsoft Azure cloud (Warsaw / Frankfurt) Implementation of the EPOD resource usage settlement model
4. 4 DATA: Integrated Complete Correct Consistent Reusable Logical Data Warehouse as a Self-Service Data Supermarket Business Model of DWH In 2017 the production environment of the Data Warehouse was modernized, the aim of which was to open the Data Warehouse for business users working in the Business Self-Service model
5. The growth dynamics of the number of DWH and BSS users  Democratization of data resulted in a dynamic increase of users with access to the Data Warehouse and the number of users logging in to the Data Warehouse  A large part of the data generated in the Business Self-Service model is distributed to end recipients in the form of Tableau reports (including Branches) 5 3-1-2017 9-1-2017 3-1-2018 9-1-2018 3-1-2019 9-1-2019 3-1-2020 9-1-2020 3-1-2021 9-1-2021 3-1-2022 9-1-2022 3-1-2023 9-1-2023 3-1-2024 0 400 800 1200 1600 DWH users No. of DWH users Average number of daily logons 2017_1 2017_2 2017_3 2017_4 2018_1 2018_2 2018_3 2018_4 2019_1 2019_2 2019_3 2019_4 2020_1 2020_2 2020_3 2020_4 2021_1 2021_2 2021_3 2021_4 2022_1 2022_2 2022_3 2022_4 2023_1 2023_2 2023_3 2023_4 2024_1 2024_2 0 4 8 12 16 20 DWH queries (excl. IFX, in millions) 3-1-2018 7-1-2018 11-1-2018 3-1-2019 7-1-2019 11-1-2019 3-1-2020 7-1-2020 11-1-2020 3-1-2021 7-1-2021 11-1-2021 3-1-2022 7-1-2022 11-1-2022 3-1-2023 7-1-2023 11-1-2023 3-1-2024 0 50 100 150 200 250 300 350 Business self-service users No. of business users with at least 1 active report No. of self-service business users 3-1-2018 6-1-2018 9-1-2018 12-1-2018 3-1-2019 6-1-2019 9-1-2019 12-1-2019 3-1-2020 6-1-2020 9-1-2020 12-1-2020 3-1-2021 6-1-2021 9-1-2021 12-1-2021 3-1-2022 6-1-2022 9-1-2022 12-1-2022 3-1-2023 6-1-2023 9-1-2023 12-1-2023 3-1-2024 6-1-2024 0 500 1000 1500 2000 2500 3000 Business self-service reports No. of active reports No. of all reports
6. 6 Reasons for the Cloud - DWH in 2023 (business use cases) (1/2) Management reporting • AKG (Analytical General Ledger) • STF (FTP) • FX settlements • Controlling Data Marts • Transactional Data Mart • SRC (Standard Risk Cost engine) • Controlling / Sales / Products / Channels: - Volumes - Profitability - Transactionality • Credit Risk - Quality of the loan portfolio - Credit workflow reports - Scoring and behavioral models - Loan portfolio reporting to PZU • BI (data source to Tableau) Regulatory reporting • Obligatory reporting (reports to the National Bank of Poland, Polish Financial Supervision Authority, EBA, BFG, Warsaw Stock Exchange) • IFRS 9 • New default • CAD (engine for capital requirements) • AMA • Market Risk (Liquidity Risk, Currency Risk, Interest Rate Risk, Other Market Risks, LCR) • CMO / STIR, SInF • FATCA/CRS • GDPR • Security (AML, KDPREVENT) Other business processes • CRM (master repository for operational and analytical CRM) • Advanced Analytics • CashBack • CJEU loan fee refunds • Bonus system for the sales network • Settlements with intermediaries and partners • „Loan Holidays” • Reports to and from BIK • Alior Bank Romania • Data Quality • Dormant accounts • Multi-product 3.000 processes 14.000 SQL scripts 4 m lines of code Over 1.100 HQ users. Over 500 logging daily Basic information: 90 TB compressed data
7. 7 Reasons for the Cloud (2/2)  Decision primarily driven by increased platform usage and additional requirements for computing power and storage  Modernization perspective to drive new analytical use cases and improvement of data time-to-market  EOL of the previous infrastructure Why upgrade?  Elastic and scalable solution allowing faster and more cost-effective deployment of additional computing power and storage through Elastic Performance on Demand (EPOD)  Strong potential for integration with other cloud services providers and adoption of new technologies evolving exclusively in cloud environments  Flexible data management and improved data accessibility for modeling  Integration with cost-effective, scalable data repositories (Azure Blob Storage accessed through Teradata Native Object Store) integrated with Teradata for cost optimization  Cloud references on a local market Why cloud?
8. 8 Data Warehouse in 2023 – physical architecture
9. 9 Data Warehouse – architecture after migration Teradata subscription Alior Bank subscription Other Alior Bank Azure Aplications Workstations BI ETL …
10. 10 Biggest risk – ensuring regulatory compliance  Banking Law in the field of outsourcing  KNF announcement regarding cloud computing  Meeting the requirements for handling incidents arising from the Act on the National Cybersecurity System  Guarantees regarding the implementation of special measures, i.e. processing of the Bank's data only in the EEA, implementation of additional encryption tools, ensuring the required system performance  Guarantees regarding the management of subcontractors in a way that meets the minimum legal requirements for the management of subcontractors in the outsourcing chain  Guarantees regarding access control to the Bank's data  Rules of conduct in the event of a security incident  Necessary audit rights for the Bank  The Supplier's liability for damage caused to customers as required by the Banking Law Selected mandatory provisions of law Selected contract areas with the Supplier
11. 11 Migration project roadmap Teradata driven Bank driven Jointly driven AUG SEP OCT NOV DEC JAN FEB KNF trial migration Platform connectivity Contract sign off Testing Stabilization Cloud environment connected Upgrade TTDB on on-prem systems Cloud platforms provisioning Trial migration PROD system PROD Final data migration Project planning Platform configuration Migration UAT system Security Tests KNF notification Implementation Thales Sending notifications NOS configuration and Migration Migration DR system Testing Final migration PROD system Testing
12. 12 Results & lessons learned – quick wins  Optimization of system operating costs, including the possibility of flexible power management (EPOD)  Reducing the processing time by an average of 45%  Data transfer time - from and to - the cloud is better than on-prem transfer time (optimizations + greater computing power) ONLY_CLOUD TRANSFER_FROM_CLOUD TRANSFER_TO_CLOUD PROD 0 500 1.000 1.500 2.000 2.500 3.000 3.500 4.000 3.666 472 2.941 1.568 192 1.481 1.457 193 1.450  Access to a wider range of analytical functionalities - Teradata Vantage Cloud and Microsoft Azure  And so… the first CRM scoring model using Teradata analytics capabilities in the cloud is there!!!  Model forecasting the customers’ propensity to log into mobile banking  Migrating cold data to Azure Blob storage and integrating them with the main environment over NOS – migrated over 72TB data / over 1.7m objects, available upon request
13. 13 Next steps  Modernize our data architecture revolving around Data Lakehouse  Leverage the business value of unstructured and semi-structured data  Mix ETL batch processing with Event Streaming and real-time capabilities to ingest data streams  Leverage more on ML/AI capabilities of the modernized data platform  Adjust our data management and data governance policy to the new requirements and implement new tools supporting data management Teradata VantageCloud Lake is one of the viable options we are considering