How End-to-End DataOps Reduce risk, Optimize Costs, and Accelerate Insights for Financial Services

In this rapidly changing, data-driven landscape, how can traditional banks compete while managing risk? How can they ensure they meet stringent privacy and security regulations like the EU’s General Data Protection Regulation (GDPR)? And, how can they position themselves to take advantage of AI and ML advances and create new customer experiences that resonate with younger, more diverse demographics?

How End-to-End DataOps Reduce risk, Optimize Costs, and Accelerate Insights for Financial Services
How End-to-End DataOps Reduce risk, Optimize Costs, and Accelerate Insights for Financial Services. Photo by Markus Spiske on Unsplash

This article discusses key market trends, common challenges and how a DataOps platform can streamline the end-to-end data supply chain to support new and emerging use cases, such as automated risk scoring and identifying and approving new, creditworthy customers.

Read on this article and learn how to quickly overcome banking’s data challenges through a modern DataOps strategy.

Table of contents

How End-to-end DataOps Reduces Risk, Optimizes Costs and Accelerates Insights for Financial Services
The Lending Landscape
The Digital Customer
Addressing Banks’ Data Challenges: The DataOps Platform
Data Sprawl and Complexity
Data Quality and 3rd Party Data
Data Governance and Regulatory Compliance
Rising Data and Vendor Costs
Key Components of Arena, Zaloni’s Modern DataOps
Platform Automated End-to-end DataOps Pipeline
Data Lineage Tracking
Zone-based Architecture
Multi-cloud & Hybrid Data Management
Collaborative Data Catalog and Self-service Consumption
Case Study: Bremer Bank Grows Lifetime Customer Value with Arena’s Golden Customer Records
Conclusion: Key Takeaways for Success

How End-to-end DataOps Reduces Risk, Optimizes Costs and Accelerates Insights for Financial Services

Learn how to quickly overcome banking’s data challenges through a modern DataOps platform

Traditional banks are at a tipping point, with a confluence of factors making the way they do business – through digital transformation – critical for their survival. In fact, nearly 75% of top executives primarily in the financial services and insurance industries said they were concerned that their companies were at risk of being disrupted or displaced by data-driven competitors. Additionally, approximately 92% of them felt that business transformation and greater agility will improve their competitiveness.

One factor causing pressure is today’s high market volatility due to the pandemic crisis, geopolitical uncertainties, and changing government policies, making it imperative for businesses, especially banks, to be more agile.

In addition, banks face significant challenges to remain competitive. For example, on the consumer-facing side, banks are losing market share to disruptors including alternative lenders and mobile payment companies – or missing out on new market segments these disruptors have tapped into. These segments include younger, tech-savvy consumers who may have non-traditional credit histories and demand a more streamlined (digital) application and approval process.

Fueling the disruptors’ growth is broader access to large volumes of data from a wide variety of sources, as well as machine learning and artificial intelligence (AI) technology, which is being aggressively leveraged for a variety of use cases, including lending/underwriting.

In this rapidly changing, data-driven landscape, how can traditional banks compete while managing risk? How can they ensure they meet stringent privacy and security regulations like the EU’s General Data Protection Regulation (GDPR)? And, how can they position themselves to take advantage of AI and ML advances and create new customer experiences that resonate with younger, more diverse demographics?

This white paper discusses key market trends, common challenges and how a DataOps platform can streamline the end-to-end data supply chain to support new and emerging use cases, such as automated risk scoring and identifying and approving new, creditworthy customers.

The Lending Landscape

The lending market is massive. The value of U.S. consumer debt, including credit cards, car loans, student loans, and mortgage debt totals around $14 trillion. Further, however large the current market, there is still enormous opportunity. According to the Federal Reserve, 31% of those who applied for credit were either denied or offered less credit than they requested.

This is where alternative lending companies are successfully using advanced analytics and machine learning to find new, creditworthy customers who may not have traditional credit history and therefore have been denied credit by traditional lenders. They’ve developed algorithms that can analyze many more sources of data to develop customer risk scores, including rental and utility payment history, short-term loans, online behaviour, bank deposits/withdrawals, mobile phone payments, social media, and more. And banks are rushing to catch up.

Beyond FinTech, some experts predict there may also be a potential threat to banks from tech giants like Google, Amazon and Alibaba, which possess huge volumes of consumer data and often use it to disrupt the status quo. The risk here is that they’ll use the data they collect to try and shake up the banking industry even further.

The Digital Customer

In addition to consumers with nontraditional credit history, traditional banks are missing out on attracting younger generations of digital native consumers who expect streamlined mobile experiences and fast, “in minutes” transactions when it comes to banking or applying for credit and loans.

Research has shown that 43% of millennials abandoned mobile banking activities because the process took too long or was too complicated. The market is expected to grow and it’s predicted that there will be 57.5 million US millennial digital banking users by the end of 2022.

Digital banking became front and centre during COVID-19, with banks needing to quickly adjust data operations for the CARES Act and Paycheck Protection Program (PPP) loan processing. The PPP program highlighted the need for streamlined DataOps to equip banks to handle sudden, large increases in complex data.

There was also demand for digital banking from retail customers during COVID-19, as branches closed or reduced hours during “shutdowns” and customer preferences shifted to online and mobile banking. A recent study states that on April 15, 2020 the daily traffic for mobile banking platforms rose 145% from the March daily average.

It’s expected that digital banking is here to stay. A recent survey by fintech company Novantas found that only 40% of respondents said they expect to return to branches post-COVID.

Addressing Banks’ Data Challenges: The DataOps Platform

Today’s challenge is to take banks’ analytics, machine learning and AI initiatives, such as automated risk scoring, beyond limited experimentation and enable them to become business drivers. Banks certainly have no shortage of data. However, successfully putting that data to work has historically presented a number of challenges. Most significant are data sprawl, data quality, lag time, data governance, and rising costs.

Data Sprawl and Complexity

As data grows and analytics use cases are implemented, data becomes sprawled across lines of business, cloud and on-premises environments and systems. According to a recent survey by Flexera, 93% of enterprises have a multi-cloud strategy and 87% have a hybrid cloud strategy, making it particularly important to have visibility and standardized governance in place across all environments along with the ability to track data lineage as data is accessed, changed or transformed.

Data Quality and 3rd Party Data

Banks spend a lot of time, money and effort on data quality. This is because having data that can be trusted is imperative for the success of any data analytics, machine learning or AI initiative. According to Gartner, organizations estimate the average cost of poor data quality at $12.8 million per year. Concern for data quality is compounded by the fact that many banks leverage multiple 3rd party data sources from vendors such as FiServ, Raymond James, Ellie Mae, and others. It’s critical that banks are able to quickly ingest, cleanse, enrich, transform, and provide access to quality, trusted data in a streamlined and simplified manner.

Data Governance and Regulatory Compliance

Banks must be able to implement and enforce strong data governance policies in order to remain compliant with multiple regulations. Data governance becomes difficult when data spans on-premises, the cloud, applications, and systems across the enterprise. In fact, more than 50% of organizations state that the lack of a standardized (enterprise) approach was the primary barrier to effective data and analytics governance. Standardizing governance across systems is essential to reducing risk and ensuring regulatory compliance.

Rising Data and Vendor Costs

Delivering a positive ROI on analytics, AI and machine learning initiatives can be difficult with rising data and vendor costs. DataOps helps to reduce these costs by automating and streamlining data pipelines to improve efficiency and accelerate analytics. In addition, to hedge against costly “rip and replace” as technologies change over time, banks should consider a DataOps platform that is extensible and easily integrates with existing and new technology so they are not locked into a particular vendor and can adapt to new use cases as they emerge.

A DataOps platform can address all of these challenges – even in complex hybrid environments – by providing:

Extensible Platform: An extensible platform that manages any data type across multiple locations and easily integrates with existing tools to control data sprawl, break down data silos, and reduce IT and vendor costs.

Collaborative Data Catalog: Collaborative data catalogue to inventory, profile, tag, annotate, enrich, share, and provide data in a way that’s automated, governed and supports collaboration across individual users and teams to improve data confidence and increase productivity.

Self-Service Marketplace: The ability for data end-users such as analysts and data scientists to easily find or “shop” for relevant data, enrich and prepare that data, and then provision it to a sandbox or analytics tool in a way that’s governed and secure, accelerating time to analytics, reducing risk, and sparing valuable IT resources.

Key Components of Arena, Zaloni’s Modern DataOps

Zaloni’s DataOps platform, Arena, gives banks the foundation they need to be more agile, test new ideas and data models, iterate fast, incorporate new technologies, and use any type of data from any source – while still complying with security and privacy regulations. Arena has helped banks achieve use cases such as:

  • Drive Revenue Growth and Assets Under Management
  • Adhere to Compliance Requirements and Legal Mandates
  • Respond More Quickly to Market Opportunities
  • AI/ML Enablement, Expert Data Curation
  • Accelerate & Automate Processes Across Systems
  • Reduce Risk Around Inquiries and Incidents
  • Facilitate Complete View of Customers, Across Systems
  • Reduce Costs IT, Analytics Efficiency

Zaloni has worked with leading financial services companies around the globe and has developed the Arena platform to meet today’s banks’ common data requirements.

Arena of Banks
Arena of Banks

Following are features and architectures that illustrate how we think about the key components of a modern DataOps platform for banks, namely:

  • Automated end-to-end data pipeline
  • Data lineage tracking
  • Zone-based governance
  • Multi-cloud data management
  • Collaborative data catalogue and self-service consumption

Platform Automated End-to-end DataOps Pipeline

It’s important to map your data pipeline and determine what parts can be automated and operationalized. In general, success at the scale of big data requires automation of key processes including data ingestion, metadata management, data privacy, and data cataloguing for self-service data users. Arena manages and automates the end-to-end data supply chain and pipelines to improve efficiency and reduce costs.

Arena DataOps Pipeline
Arena DataOps Pipeline

Data Lineage Tracking

As your data moves through a typical workflow – ingestion, data quality checks, transformations, masking, tokenization, etc. – it is imperative to track the provenance or lineage of the data in order to understand its quality. You want to be able to drill down to see what steps the data went through and what actions were taken – and do this across a hybrid or multi-cloud environment. Below is an example of the level of detail provided by Arena to improve visibility, confidence, and security.

Zone-based Architecture

Based on hundreds of data platform implementations, Zaloni developed a best-practice zone-based governance architecture, EndZone Governance™, that works across environments – on-premise, cloud, and multi-cloud. Creating “raw,” “trusted,” “refined,” and “sandbox” zones are a practical way to manage data quality, define and secure data access, and enable you to categorize data by the different stages of its lifecycle. Raw data can be kept for historical records; once transformed, data is validated, catalogued and assigned metadata as it moves to subsequent “refined” and “trusted” zones. Role-based access provides governance controls over who has access to which zone.

Multi-cloud & Hybrid Data Management

Most banks are interested in moving to the cloud for both storage and compute benefits. A hybrid environment gives you the best of both worlds by enabling you to keep critical data on-premise while also leveraging cheap storage and the on-demand nature of the cloud to run analytics models. A modern DataOps platform also allows you to leverage multiple cloud environments, providing agility and avoiding vendor lock-in. The challenge is managing data spread across environments and making it accessible for end-users – it’s key to have a robust DataOps platform that can provide this visibility and control. Below is an example of a multi-cloud architecture using Arena.

Collaborative Data Catalog and Self-service Consumption

Being a data-driven organization requires making data more broadly available to more business users. A platform with strong data governance allows you to provide self-service access to data for analytics via a rich data catalogue and global search that allows business users to search across data, zones, projects, workflows, data quality rules, transformations and other related content. Arena’s data catalogue promotes collaboration and productivity by allowing data users to annotate, tag, and create custom metadata, as well as share relevant data to users and teams. Once data is found, users can quickly and easily provision data to an analytics tool or sandbox environment.

Case Study: Bremer Bank Grows Lifetime Customer Value with Arena’s Golden Customer Records

Regional financial services company, Bremer Bank, provides financial services offerings including banking, insurance and investment management, and wanted to improve customer experience to increase customer satisfaction and uncover cross-sell and upsell opportunities to drive top-line growth.

Challenge: Bremer Bank was modernizing its technology systems and wanted to centralize siloed data across their lines of business. Due to their organizational focus on customer experience, they needed to integrate all customer data sets to create trusted golden records of their customers that could be leveraged and synchronized across their five lines of business. The company faced challenges matching and mastering data from various systems and sources including internal, external and 3rd party data sources like EPIC, Raymond James and internal CRM systems. Many of the data sets lacked unique identifiers and had critical fields missing creating integration and enrichment challenges.

Solution: Arena allowed the company to architect a cloud-based data lake in Azure. The lake was hydrated with an initial set of data sources to understand data set structures, requirements and what the end result or golden record should be, then the solution was built using Arena’s data mastering to put customer golden record creation into production. Arena’s data mastering is powered by a machine learning engine and data mastering continuously improves over time as new data sets are added. Using Arena, the company was able to build a data science workbench to help data scientists to uncover new data monetization opportunities.

Results: Arena enabled Bremer Bank to build golden customer records that are leveraged by sales, marketing and customer success to improve customer experience, provide personalized marketing offers, and increase revenue by uncovering cross-sell and upsell opportunities. Arena saved 8 hours/ day of data engineering work, reducing costs, and improving efficiency. The data science group was able to uncover new data monetization opportunities through the workbench capabilities within the platform.

Conclusion: Key Takeaways for Success

The banking landscape has changed and will continue to do so. However, no matter how consumer preferences, technologies, artificial intelligence and machine learning, cloud offerings or modelling techniques evolve, staying in the game requires banks to have comprehensive oversight of their data supply chain. This is core to whatever the future might bring. “Modernizing” your data infrastructure means architecting it in such a way to enable both tight control over data quality and governance and the flexibility to adapt over time. As you move forward in your digital transformation journey, we urge you to keep these key considerations top of mind:

  • Ensure data quality by applying active metadata and tracking data lineage
  • Leverage operational efficiencies by automating and operationalizing processes throughout the data supply chain
  • Mitigate risk with strong data governance, security and role-based access controls
  • Reduce costs by streamlining the data supply chain
  • Implement an extensible platform that provides use case agility and doesn’t lock you into technologies or vendors
  • Insist on transparency for machine learning models by using good data management hygiene to avoid unknown, “baked in” biases that can make models less accurate

Zaloni’s Arena platform enables DataOps success by streamlining the data supply chain and providing data unification, discovery, governance, active metadata management, collaboration, mastering, and self-service provisioning, in one unified extensible platform. Arena secures data pipelines to enable better, faster analytics, reduce the burden on IT, and lower data costs. We work with banks around the globe to help them transform their businesses through data.

Source: Zaloni

Thomas Apel Published by Thomas Apel

, a dynamic and self-motivated information technology architect, with a thorough knowledge of all facets pertaining to system and network infrastructure design, implementation and administration. I enjoy the technical writing process and answering readers' comments included.