Transcription
DatabricksCustomer Story:BancoHipotecario
Data and ML creates asecure and personalized 21st centurybanking experienceBanco Hipotecario boosts customer lifetime value with DatabricksThe best part is that wecalculate that by usingDatabricks, we’ve savedaround 90% of the costof having a Spark clusterin our datacenter.Banco Hipotecario, a leading Argentinian commercial bank, is ona mission to reduce customer risk and provide an experience thatwill increase retention and loyalty. Key to this strategy is to leveragemachine learning to deliver new insights and services that willdelight customers and create upsell opportunities. With a legacyanalytics system that was rigid and complex to scale, they turned toDatabricks to unify data science, engineering, and analytics. As aresult, they were able to significantly increase customer acquisitionand cross-sells while lowering the cost for acquisition, greatlyimpacting overall customer retention and profitability.INDUSTRYFinancial servicesSOLUTION Customer 360 Personalized experience Recommendation engineTECHNICAL USE CASE Data ingest and ETL Machine learning SQL analyticsMAT I AS J. STA N I S L AVS K YHead of BI and Advanced Analytics atBanco Hipotecario90%INCREASE INCROSS-SELL OF NEW PRODUCTS,35%REDUCTION IN COSTOF ACQUISITION, ALLOWING THEMIMPROVING CUSTOMER RETENTIONTO DO UPSELLS AND CROSS-SELLSAND PROFITABILITYTO ALL THEIR CUSTOMERSDATABRICKS CUSTOMER STORY: BANCO HIPOTECARIO3
Legacy analytics tools are slow, rigid, andimpossible to scaleA unified platform powers the data lakeand easy collaborationBanco Hipotecario set forth to increase customerto model training. They had a traditional analytics toolBanco Hipotecario turned to Databricks to modernizeacquisition by reducing risk and improving the customerrunning on SQL databases and their data warehouse wastheir data warehouse environment, improve cross-teamexperience. With data analytics and machine learningnot well organized and complicated. Their data scientistscollaboration, and drive data science innovation. Fullyanchoring their strategy, they hoped to influence a rangewere slowed by disparate data sources and the inabilitymanaged in Microsoft Azure, they were able to easilyof use cases from fraud detection and risk analysis toto scale machine learning due to single-node limitations.and reliably ingest massive volumes of data, spinning upserving product recommendations to drive, up-sell andcross-sell opportunities and forecasting sales.Banco Hipotecario knew that if they wanted to shiftwith the times, as they’d done for decades, they’d needLike most traditional companies, Banco Hipotecarioto overhaul their existing tech stack and empower theirfaced a number of the challenges that often come alongdata team with the tools to be productive and innovate.with outdated technology and processes: disorganizedor inaccurate data; the inability to innovate and scale;resource intensive workflows, -- the list goes on.“In order to execute on our data analytics strategy,new technologies were needed in order toimprove data engineering and boost data scienceExisting data teams, including the data engineers andproductivity,” said Daniel Sanchez, Enterprise Dataanalysts to data scientists, also found it difficult, andArchitect at Banco Hipotecario.”The first stepssometimes impossible to collaborate and leverage thewe took were to move to a cloud-based data lakedata for their specific needs from business intelligencewhich led us to Azure Databricks and Delta Lake.”their whole infrastructure in ninety days: Delta Lake, ETLpipelines, integration with Azure’s Data Factory, MLOps andDataOPs, drastically improving workflows for downstreamanalytics and machine learning.Today they no longer waste time managing and configuringclusters. With Databricks’ automated cluster managementcapabilities, they are able to scale clusters on-demand tosupport large workloads.Delta Lake has been especially useful in bringingreliability and performance to Banco Hipotecario’sdata lake environment. With Delta Lake, they are nowable to build reliable and performant ETL pipelineslike never before.“We work at a bank and our transactions can be rejected,At the same time, Data scientists were finally able toor turned down a couple of days after the transaction hascollaborate thanks to interactive notebooks, meaningbeen completed,” explained Sanchez. “In a traditional datafaster builds, training, and deployment. And MLFlowlake, when you try to do an upsert, you need to delete a filestreamlined the ML lifecycle and removed the overrelianceand rewrite it. Now with Delta Lake, we just make upsertson data engineering.and let the engine work to make all the necessary changeson the files.”“Databricks gives our data scientists the meansto easily create our own experiments and deploy themMeanwhile, performing SQL analytics on Databricks hasto production in weeks, rather than months,” said Miguelhelped them to do all of the data exploration, cleansing,Villalba, Head of Data Engineering and Data Science.and generation of datasets in order to create models,enabling the team to deploy their first model within thatfirst 3 months, and the second model generated was offto the races in just two weeks.DATABRICKS CUSTOMER STORY: BANCO HIPOTECARIO5
An efficient team maximizes customeracquisition and retentionSince moving to Databricks, the data team at BancoThe results of data unification, and markedlyHipotecario could not be happier as it has unified themimproved collaboration and autonomy cannot beacross functions in an integrated fashion. From anoverstated. Since deploying Databricks, Bancoarchitectural perspective, Azure Databricks has greatlyHipotecario has increased their cross-sell intosimplified their entire data analytics ecosystem and end-new products by a whopping 90%, while machineto-end data workflow.learning has reduced the cost of customer‘Databricks has been a game changer,” expressedStanislavsky. “It has empowered our entire organizationwith the same technologies that the top companies in theworld are using to innovate.”acquisition by 35%.Stanislavsky concluded, “The knowledge and capabilitiesthat we’ve gained has been simply awesome. From usingPySpark to the integration with Azure, we are a completelydifferent organization now.”DATABRICKS CUSTOMER STORY: BANCO HIPOTECARIODatabricks gives ourdata scientists themeans to easily createour own experimentsand deploy them toproduction in weeks,rather than months.M AT I AS J. STANI S L AVS KYHead of BI and Advanced Analytics atBanco Hipotecario7
About DatabricksDatabricks is the data and AI company. Thousands of organizations worldwide —including Showtime, Shell, Conde Nast andRegeneron — rely on Databricks’ open and unified platform for data engineering, machine learning and analytics. Databricksis venture-backed and headquartered in San Francisco with offices around the globe. Founded by the original creators ofApache Spark , Delta Lake and MLflow, Databricks is on a mission to help data teams solve the world’s toughest problems.To learn more, follow Databricks on Twitter, LinkedIn and Facebook.E VA LUATE DATAB R IC KS FOR YOU R S EL FS TA R T Y O U R F R E E T R I A LContact us for a personalized demo databricks.com/contact Databricks Inc. 2020. All rights reserved. Apache, Apache Spark, Spark and the Spark logo are trademarks of the Apache Software Foundation. Privacy Policy Terms of Use
Legacy analytics tools are slow, rigid, and impossible to scale Banco Hipotecario set forth to increase customer acquisition by reducing risk and improving the customer experience. With data analytics and machine learning anchoring their strategy, they hoped to influence a range of use cases from fraud detection and risk analysis to
