Case Studies
The Work
Each of these represents a build — not an optimization of something that already existed. Context, problem, action, outcome.
Full Lifecycle Data Organization Build
Context
Jenius Bank was a new digital bank being built from scratch inside SMBC. There was no data team, no cloud environment, no governance, and no analytics culture. The business itself was still being formed.
The Problem
Most organizations talk about building a data culture. Jenius Bank had to build one from nothing, while the business was also being built around it. Everything needed to be created — infrastructure, pipelines, visualizations, governance — with no legacy to inherit and no existing team.
Fraud Analytics & AI Lab Redirection
Context
USAA is one of the largest financial services organizations in the U.S., serving military members and their families. The fraud and enterprise analytics division handled detection, modeling, and reporting across the organization. The AI Lab was a separate initiative focused on advanced analytics.
The Problem
Fraud signal detection was fragmented across channels, creating noise and missed connections. Separately, the AI Lab had significant investment but was oriented toward academic research without clear ties to business outcomes or revenue generation.
Rapid Analytics Impact
Context
Ally Bank is a leading digital financial services company. The analytics function needed acceleration — stronger capabilities, faster delivery, and better alignment with business strategy.
The Problem
The analytics team needed to deliver faster results while building longer-term capability. The challenge was creating immediate value without sacrificing foundational work that would enable sustained growth.
Deep Dive
Getting the Business in the Room
The hardest governance problem is not technical. It is organizational. Different teams used the same terms to describe different things. No one had documented what a customer was, what an account was, or who was responsible for either definition.
Getting in the Room
Executives do not naturally volunteer to document data definitions. The case I made was direct: if you cannot define your data, you cannot trust your reports. If you cannot trust your reports, you cannot run your business.
Building Ownership
We built the data owner structure, assigning accountability at the business level rather than the technology level. For the first time, the people who created the data were responsible for its definition.
Domain & Data Mart Strategy
We designed the domain and data mart strategy that would organize data around business concepts rather than system outputs. Business-centric data architecture: the right model for any institution.
Architecture for What Comes Next
We built the roadmap to deliver the strategy. The documentation, the ownership model, and the architecture are in place. This is the foundation the next team will build on.