The Challenge: Data everywhere, and going nowhere fast

Like many credit unions, All In Credit Union was drowning in data and needed insights faster and more reliably. Reports, tools, and data sources were scattered across multiple systems, owned by various teams, with no single source of truth.

All In’s SVP of Sales and Lending, Todd Peeples, and his team aim to be data-driven and conduct thorough due diligence before taking action on strategies to drive business performance. However, presumably simple requests—like breaking down auto loans by credit tier and performance to move forward on a new initiative—could take two weeks, involve 3–4 staff members, including multiple systems that contained and/or analyzed the data, and still come back incomplete. Most requests had to go back and forth several times, stretching the process even longer. The inefficiency wasn’t just frustrating—it slowed and even stalled decision-making.

Todd evaluated the layers of touchpoints needed to access and analyze data as well.

“Reporting and analysis are often a resource-intensive process. Reliability in the insights and output is also a challenge because I may have something in my mind that I want to see; however, someone on the data team may use other terms or language and interpret the request differently.”

Not just any AI will do

Peeples is no stranger to the value AI could bring to many parts of the organization, having been an early adopter of advanced machine learning in his loan decisioning processes. However, his team was very concerned about security and governance when it came specifically to generative AI. They wanted to harness the power, but not expose themselves or their data to risks.

Building a solution in-house that would compete with the big banks and fintechs was costly and resource-prohibitive. BI tools that bolted on AI still had the same underlying issues with extracting and analyzing data. Popular large language models (LLMs) are too generic and don’t have industry context or access to the most accurate data.

Enter LuLu, a generative AI (GenAI) lending intelligence companion that was developed by Zest AI, a long-time partner of All In for their AI-automated underwriting solution. LuLu is GenAI that is purpose-built for lenders from the ground up—sourcing and maintaining abundant, curated, relevant data in a secure and customized way to support lending organizations. The team was able to access industry data sources such as NCUA call reports and HMDA data, as well as updated macroeconomic data, proprietary data, and more, in one simple hub.

Previously, using a discombobulated array of tools for regular reporting needs, including BI, platforms, and spreadsheets, could now be consolidated by 50% as the data sources continue to expand. Todd and his team could have instant, direct access to answers, saving over 80% of the total time and resources in reporting.

“What’s nice about LuLu is it’s been purpose-built AI from the get-go. That’s a big advantage when it comes to extracting actionable insights. To have a solution where anyone can use everyday language to dive into the data—and get immediate, reliable results without waiting—that is huge.”

Accelerating efficiency, accuracy, and adoption with LuLu

All In isn’t just getting data faster—they’re getting smarter recommendations that lead to real change. For example, Todd was able to ask a simple question on how they could optimize their portfolio performance, and LuLu identified opportunities and quantified the potential impact.

“LuLu gave me eight or nine specific suggestions—with percentages tied to potential lift, like 3% here or 4% there. We’re not doing all of them, but some we’re acting on immediately.”

Getting a team to adopt new technology, especially analytics tools, can be difficult due to the complexity and general change management challenges. With LuLu, the process was easy.

“The onboarding process was smooth, and the team did a great job showing us how to use the tool effectively. But honestly, even without training, you can pick it up right away. If you can use Google, you can use LuLu—it’s that easy.”

With the ability to surface data and analysis proactively and directly, decision-makers throughout the organization can trust the accuracy of the information. This confidence enhances efficiency and agility, enabling them to take action quickly. Peeples estimates that this ability saves over 90% of the time it takes from analysis to moving forward on a plan.

The future of credit unions is bright with GenAI

Todd believes that GenAI is a competitive advantage that credit unions need. Credit unions hold just over 10% of the lending market, yet their influence is immense for the members they serve while also strengthening their communities. Leveraging AI helps credit unions compete more effectively while staying true to their mission of people helping people.

“Credit Unions need technologies like this to help us grow and move forward. We’d rather find partners who are innovative and forward-thinking like Zest AI. Most credit unions don’t have internal resources to build this themselves. LuLu is built for us—so we can compete and win.”