Founded in 1858 and headquartered in Cincinnati, Fifth Third Bank stands among the nation’s most established financial institutions. Today, the bank’s modernization journey is powered by data and AI — not as buzzwords, but as practical levers embedded deep within its operating model.

In this third and final installment of the three-part series, Shanna Anderson, Sr. Director, Customer Data and Insights at Fifth Third Bank, joins Or Zabludowski, CEO of Flexor, to share her actionable playbook for data leaders in financial services — one built on aligning AI with business impact, scaling automation across unstructured environments, and fostering trusted, self-service access to data.

The first part of this series unpacked the bank’s evolving ModelOps journey and how it is laying a foundation for trusted, explainable AI. The second installment examined how the bank ensures that every AI initiative has a clear, measurable impact while keeping people at the center of decision-making.

Bridging the paper-to-AI Gap in a legacy industry

Anderson begins by highlighting a central challenge faced by many long-standing banks: The use of paper. “Banking is still very much a paper business in some aspects,” she explains, noting that despite digitization efforts, customers continue to write checks and submit physical documents for loans and other services.

This creates both friction and opportunity. Her team actively explores use cases where AI and generative AI can be deployed to streamline paper-heavy, manual processes.

Anderson sees a broad landscape of untapped value, not just in physical paperwork, but in unstructured data such as lengthy documents, customer correspondence, and policy manuals. By leveraging AI to extract, structure, and process this data, the bank is beginning to unlock new forms of efficiency and insight.

Unlocking insights from unstructured information

The conversation turns to the transformational potential of turning word-heavy documents into machine-readable, structured data. Anderson acknowledges this shift is already underway: “So much can be unlocked, while also shortening the time it takes for employees within a company to glean insights.”

Rather than parsing 100-page documents manually, employees can now query structured outputs derived from text, accelerating decision-making and reducing the knowledge gap. Whether it is underwriting, risk evaluation, or customer interaction history, AI offers a new layer of augmentation for human intelligence.

This ability to scale insight, especially from previously underutilized data, paves the way for more advanced capabilities like prescriptive analytics and historical trend prediction.

The AI playbook: Business first, automation second

When asked to outline her top recommendations for data leaders in financial services, Anderson shares a playbook built on business empathy, self-service enablement, and a commitment to data trust:

1. Fuse business and tech goals: Anderson emphasizes the value of a deep partnership between business leaders and data teams. “It doesn’t feel like it’s us versus them,” she says. “Think about the pain points that the business teams feel and how my teams can make that better?” This mindset allows data teams to focus not only on delivery but on delivering what truly matters.
2. Modern data strategy with embedded access: A standout capability Anderson highlights is Fifth Third’s internal data marketplace — a curated repository of trusted data products. Designed to prevent the enterprise data office from becoming a bottleneck, the marketplace automates access, displays data lineage, and shows both business and tech ownership. Most critically, it removes technical barriers for non-engineers. “They don’t have to feel like they have to understand ETLs or SQL to be able to work with data.”
3. Automate for accessibility: Anderson believes automation is the enabler of federation. For business users, the idea of owning or managing data can be daunting. “When it’s automated and they’re able to get to it without having to do the lift, that’s a big win.”
4. Lay a strong data foundation for AI: She underscores the need to rebuild legacy processes for modern AI consumption. “If we don’t have the right data foundation set up, it either won’t be trusted or it will fail.” Her team spends significant time re-documenting and modernizing previously opaque workflows to ensure resiliency, transparency, and business confidence in AI-driven decisions.

Avoiding the tech-for-tech’s-sake trap

As a final caution, Anderson warns against pursuing technology without business alignment. “Watch the tech-for-tech,” she advises. While keeping platforms healthy is essential, innovation must always be tied to value creation and strategic business outcomes. Data teams should be able to articulate not just what they’re building, but why and what impact it will have.

Being able to trace outcomes back to development efforts helps reinforce trust between data and business teams. “That goes a long way with teams and partnerships,” she concludes.

CDO Magazine appreciates Shanna Anderson for sharing her insights with our global community.