Increasing Efficiency for Admin Work and Processes
July 2026 – Present
Transforming the TANF Data Portal admin experience from a complex, table-heavy tool into a clearer and more flexible, customized experience that helps government staff users, monitor system activity and complete everyday tasks with fewer steps.
Role
Product Designer
Tools
Figma, MaterialUI, OpenAI, GitHub
Client
The Administration for Children and Families (ACF) DataTeam
Industry
Civic Tech
Product Design
React Web App
Admin Tools
Systems Thinking
Cross-Functional Collab
A complete transformation of the admin experience, from framework to functionality.
The designs shown in this case study have been modified to protect sensitive application data and user privacy
Background and motivation
Our customer, the TANF Data Portal Data Team, relied on an antiquated system for their admin tasks, from managing users to monitoring system activity and troubleshooting data issues. But as their needs grew, the framework became increasingly limiting for both the people using it and the developers maintaining it. The core functionality was there, but the framework needed to be moved.
The opportunity
It was time to move the experience into React and rethink, not only the visual layout, but the way work has been getting done from the ground up. This was not just simply a lift and shift from one framework to another. I focused on making common tasks easier to complete, important system activity easier to understand and next steps clearer, while preserving access to the detailed information administrators still rely on while introducing new features to enhance their work.
Impact and results
This work is still ongoing between design, development and testing
Before implementation began, the redesign gave the team a clear, validated direction for the new admin experience. Our customers, who are also the primary users of the tool, were involved throughout the process and responded very positively to the direction. Working closely with them and the development team helped turn technical requirements and existing pain points into an experience that felt useful, realistic to build and worth getting excited about using.
Reduced steps in common workflows such as user request approvals
Improved visibility into what needs attention, action or investigation
Turned raw feedback into clearer insights with generative summaries, supporting stats and context about where feedback came from
Validated with users and development before implementation
Designed to scale for additional administrators and collaborative workflows
Process, Approach & Methods
Requirements Gathering
Legacy Audit
Figma
Systems Thinking
Stakeholder Collaboration
Dev Collaboration & Handoff
Requirements gathering and technical constraints
With every project, I start by understanding how the current experience is actually being used. Because our customer is also the primary user of the admin tool, I was able to ask about their most common tasks and observe how they navigated tables, interpreted system information and made decisions.
Those observations surfaced pain points that were easy to miss from the interface alone. Simple tasks often required scrolling through rows of data, interpreting unclear statuses or return values and clicking into additional records just to understand what had happened.
That led to one of the biggest opportunities for the redesign: provide more context at the point where decisions are being made. Instead of only showing raw data, we could use clearer labels, summaries and simple system logic to help administrators understand what needed attention and what action to take next.
From the development and technical implementation side of the house, I needed to consider:
The established USWDS design system that was is being used for the user facing application and utilizing the established library and balancing the trade offs with creating more custom components for our admin users
Material UI iconography
Finally I translated what we learned into a small set of guiding user stories and requirements that went beyond simply “redesigning the existing pages.” Here are a few main ones:
Samples of the Django admin experience
Primary User Story 1: Manage access with fewer steps
As a system administrator, I want to review and act on user access requests efficiently so that I can manage permissions without unnecessary steps or technical interpretation.
Primary User Story 2: Quickly understand what needs attention
As an administrator, I want a clear view of system activity and issues that need attention so that I can quickly understand what is happening and where I need to take action.
Primary User Story 3: Understand Feedback in Context
As an administrator, I want context around user feedback so that I can understand what the data means and identify where follow-up may be needed.
Sample sketches of the initial phases of generating ideas and considerations
More than just 1:1 translation of existing functionality. Thinking big by creating little features.
Rather than simply recreating existing functionality, I explored small features that could make the admin experience more useful. These included clearer error states, creating new tickets, bringing related conversations together and helping admins stay informed.
Components exploring error states, ticket creation, consolidated conversations and notifications that keep admins informed.
Development feasibility checks, handoff and iterative loops
I tend to design the art of the possibility and what near final states could look like and from there, scale back iterations to more digestible pieces. Sometimes this involved creating near term designs and longer term thinking. Generative AI summaries was one of these examples. I created one design that could aggregate the data we have into line tables to display counts of information and where the feedback was coming from, and another to create generative summaries of user feedback based on satisfaction selections.
Samples of Figma annotations that I noted for the development team
Using AI to surface patterns in user feedback
I intentionally incorporated AI to help analyze user feedback at scale, turning individual submissions into actionable insights. By aggregating feedback across submissions, I explored how AI could identify positive and negative sentiment, surface recurring themes and highlight where feedback was submitted. This approach could help teams quickly identify emerging pain points, recognize what was working well and prioritize opportunities for improvement without manually reviewing every submission.
Up Next: Rethinking Data Tables as Reusable Components
I'm currently exploring how to build reusable data tables that save development time while still meeting the needs of different workflows. As I dig deeper, I'm finding that each table has its own quirks and the data I have access to doesn't always show every possible scenario. I'm continuing to work with developers and refine the designs to figure out what can be shared across tables and what needs to be customized.
Rethinking how data is presented in tables
Measuring success through usability studies
I've outlined how I would validate key design decisions and measure their impact through usability testing, user feedback and product metrics.
Design hypothesis
Testing method
Success metric
Simplified navigation will help users find information faster
Usability testing
Time on task, task success
Streamlined workflows will reduce unnecessary steps
Comparative usability testing
Completion time, error rate
Improved information hierarchy will make content easier to understand
Usability testing and interviews
Comprehension accuracy, perceived ease
New features will better support user needs
Product analytics and feedback
Adoption rate, satisfaction
Reflection
This project has been a great opportunity to showcase my design thinking and problem-solving skills to my team and our government partners who are extremely excited about this product. One of the biggest challenges has been balancing new design ideas with existing frameworks and design systems. While the team was excited about the designs, we also needed to consider what was practical to implement. The development team recommended starting with the core USWDS components and rolling out improvements in phases.
