Project Case Study
Building the Gates Foundation's enterprise data platform
I helped scope, design, and launch a platform that balances the needs of data scientists and program officers with enterprise data governance objectives and AI readiness.
Context
The Bill & Melinda Gates Foundation had long operated with a decentralized, bring-your-own-tools approach to data — workable at smaller scale, but increasingly limiting as the foundation grew its appetite for data-driven investment strategy, cross-team collaboration, and AI-enabled analysis. Data lived in disparate systems across teams, external collaboration required slow manual workarounds, and there was no central place to find, catalog, or govern the foundation's growing portfolio of datasets.
A previous internal data repository had addressed storage and cataloging but never fully delivered. Adoption was hampered by overly burdensome metadata requirements and a lack of integration with the foundation’s other systems, processes, and policies. The technology landscape had also evolved substantially since that technology was chosen. In 2024, leadership commissioned a new Enterprise Data Platform (EDP) designed from the ground up to be modular, secure, and future-ready.
Team & Role
I wore multiple hats within a large team across multiple phases of work, partnering most closely with a business systems analyst who served as Agile Product Owner, one contract designer, engineers, data architects, business stakeholders, and IT leadership.
- As a UX Researcher & Strategist, I led discovery research across data scientists, analysts, program leaders, legal, information security, and external collaborators – synthesizing findings into personas, journey maps, and pain point frameworks that shaped platform scope, business case, and MVP priorities. I facilitated team prioritization sessions, a CIO-level data governance alignment workshop, and pilot testing with early adopters.
- As an Information Architect, I defined the conceptual model and metadata schema underlying the EDP Data Portal – a custom web app we built as the “front door” of the platform, providing user-friendly intake, secure transfer, and data catalog search functionality. I facilitated working sessions with legal and enterprise data stakeholders to balance governance needs with usability, informed by lessons from the prior data storage solution where overly burdensome metadata requirements had contributed to limited adoption. I also defined naming conventions for our Databricks Unity Catalog (underlying governance layer and primary interface for more technical data science users).
- As Product Designer I led design for the custom Data Portal web application, collaborating with 1 contract designer on both the big picture and fine details – including a branded design system built on Microsoft Fluent, intake and upload workflows, cataloging features, dataset search, and role-based access controls. I adapted the foundation's recently rebranded visual identity for complex web application contexts, conducted design QA, identified accessibility and responsive issues for remediation, navigated scope negotiation when development timelines slipped, and provided ongoing design direction to the contract designer.
Obviously, there's a lot to unpack from this two year project – deep dive case study and process visuals coming soon!
Outcomes
After a successful pilot phase, the platform launched in August 2025.
Within first 6 months, 37 datasets had been cataloged and ingested across 22 foundation teams (71% of team adoption target, ahead of plan).
Scaling continues in 2026.
Designing streamlined dashboards for managing an $8bn philanthropic budget
Building a virtual clinic to connect patients & clinicians 24/7