Transforming a statewide product and the design organization behind it.
I joined a high-pressure modernization where design was struggling to reconcile worker needs, state policy, stakeholder expectations and technical constraints. I rebuilt how the team worked while reshaping the product foundation for scale.
CWS-CARES was operating under critical timelines, but the design organization was caught between competing realities. Workers needed an experience that reflected how they actually managed complex cases. State policy defined what the system had to support. Stakeholders brought another layer of priorities and expectations. Engineering had technical constraints that were not consistently understood by design.
The result was significant rework. Approved designs did not always translate into what was ultimately built. The product lacked a coherent information architecture, reusable patterns and a scalable component foundation. Workers were becoming frustrated, stakeholders were losing confidence and the existing approach could not scale across a statewide platform.
I met separately with designers, engineers, developers, leadership, workers and stakeholders to understand how each group experienced the problem. The failures were not isolated. Team structure, product structure, delivery and decision-making were reinforcing one another.
I reorganized design responsibilities around capabilities and experience. I established regular working cadences between design and engineering so technical constraints, customization possibilities and out-of-scope requests were understood before solutions reached approval.
I also led workshops with workers to understand their needs and frustration while helping them understand policy and technology constraints. The goal was not to let one group win. It was to create a product direction everyone could understand and execute.
Before · sequential handoffs
After · shared decisions
The lack of IA was not simply a navigation problem. It affected how workers understood cases, how information was collected, how data moved through the application and how future capabilities could scale.
I worked directly on the IA and discovery alongside the team. The new structure allowed the experience to move away from fragmented information toward a more complete case context. Workers could connect related cases, keep relevant information open, navigate nested information more logically and access a broader global view of a case.
I introduced the design system as a solution to a scalability problem. Instead of design producing a library and handing it to development, designers and engineers built the component foundation collaboratively. That gave engineering a voice in technical behavior while giving design a reusable system for maintaining experience consistency.
Beyond the metrics, the program had a stronger design-engineering partnership, more consistent product decisions, a reusable product foundation and a better ability to scale across workstreams.
The outcome was not simply a redesigned application. The work left behind a stronger design organization, operating model, information architecture, cross-functional cadence and reusable product system.
Visuals are reconstructed and anonymized to protect client confidentiality.
Designing responsible AI for high-stakes child welfare decisions.
Pennsylvania DHS · Responsible AI · Human-in-the-loop · Statewide modernization