How Fifth Third Bank Democratizes Data Access via a Data Mesh with Alation and Snowflake

By Talo Thomson

Published on 2022年6月7日

How Fifth Third Bank Democratizes Data Access via a Data Mesh with Alation and Snowflake

Today, moving money between accounts takes a minute of attention and a few swipes of the finger. iBanking has become so easy and advanced that we take it for granted. Yet we’re still in the very early days of digital banking — and what it can do.

Fifth Third is helping redefine that future with a commitment to better understanding and serving their customers. This nationally chartered bank is headquartered in Ohio, and today is among the largest money managers in the Midwest. Most people don’t know Fifth Third was the inventor of the ATM. And today, the organization remains laser-focused on innovations that support their customers. The key to that innovation is data.

Yet Fifth Third’s vast data environment presents a number of challenges. A small group of data leaders faced an explosion in both the need and demand for data — and a lack of the structure to support it. To meet that demand, they’ve focused on delivering data self-service across the organization. This means adopting data mesh and data-as-a-product principles, and transforming their infrastructure with the help of a data catalog and warehouse, revolutionizing their data culture along the way.

Fifth Third’s pain-points listed out as keywords clusters.

The Problem: The Data Challenges

The data challenges at Fifth Third will sound familiar to anyone working in an enterprise data landscape. A small number of data scientists and engineers faced a clamoring crowd of business users who needed access to data. To meet that growing demand, they decided to make everyone a data citizen. Today, they bridge the gap between the experts with data and everyone who needs to use data with a self-service environment; in other words, they’ve democratized data, supported by a system of record, with clear, authoritative sources and labels.

“This is an economies of scale problem,” points out Kayleigh Lavorini, product owner – data strategy, Fifth Third Bank. “We’re never going to be able to hire enough data engineers, data scientists, and cloud architects to support the growth that we want to achieve. But we do have a ton of really amazing, really skilled, really smart and competent people that we can upskill, invest in, give the right technology, give the right rails, and turn them into the citizen data engineers.”

The principles of a federated data mesh, Fifth Third wagered, would enable the company to democratize data and create a self-service environment in which data is treated as a product and more people can independently access, understand, and use that data as needed.

The Solution

Fifth Third had already proven that data intelligence was key to delivering data to the right people. Their data catalog project proved that everyone could find, access, and know how to use data. To tackle this bigger problem of bringing data to everyone, Kayleigh and her team created a data strategy that transformed the rest of the data stack.

They took their centralized architecture and are creating a decentralized, cloud-native and domain-centric data environment. The Snowflake Data Cloud serves as their central repository for data and analytics, and their Alation data catalog now provides the metadata management capabilities to all data citizens.

“We’ve got this central, very tightly coupled, architecturally central repository managed by a single team,” Lavorini reveals. “And it provides the grounds for all of the analytics at the bank. Thousands of employees can access hundreds of reports and data products. They hit that warehouse every single day.”

Yet process and technology are just half of the equation. “What we believe is that this is much more than a technical change,” Lavorini continues. “It’s shifting the beliefs and the thought process around data, and it’s believing in and creating a data culture.” Changing the habits around people and data today sets the foundation for an innovative tomorrow.

“Through that, we imagine a Fifth Third where data and insights are created by all employees, powered by us as a central data services organization, and made available and shareable and discoverable by anyone in the company as a product or an asset,” she concludes.

The Result

That vision — where data and insights are created by everyone, powered by a central team, and shared across the enterprise — is already bearing fruit. Today, Alation and Snowflake’s joint data governance solution allows Fifth Third Bank to:

  • Trust. People can identify the right source for a data element so they know they’re pulling data from the right place.

  • Guide. The data catalog inventories data sources and describes data elements with guidance on intended use in a single glossary and dictionary.

  • Understand. Transparency into data lineage and the data supply chain empowers people to know when data was first created, and who to ask for answers.

At Fifth Third, Snowflake’s data warehouse is the one-stop shop to store and transform cloud data and data products. Alation adds a semantic layer, with metadata labels, to guide search & discovery, governance, and enable stewards to curate the best data (and make it visible to all).

Conclusion

Kayleigh and her team at Fifth Third Bank have built a resilient tech platform, supported by Alation and Snowflake, that empowers a wide range of user to self-serve and use data with confidence. This foundation is critical to innovating with data in the years to come.“Fifth Third puts the customer at the center of everything we do,” Kayleigh elaborates. “Maturing our data strategy helps to accelerate our value to the customer.”

Curious to learn more about how building a data mesh strategy can support self-service analytics and data governance?

Join Kayleigh Lavorini, product owner of data strategy, Fifth Third Bank, for her session, Building a Data Mesh: Fifth Third Bank’s Data Strategy Journey at Snowflake Summit on June 14, 2022 at 3:30 p.m PT.

    Contents
  • The Problem: The Data Challenges
  • The Solution
  • The Result
  • Conclusion
Tagged with