Chat Public Benefits Assistant (ChatPBA)

What do we want to achieve?

- Create an AI chat for public benefit navigators to get quick and accurate answers to questions about public benefits to help their clients.
- Design a UI design that is intuitive and easy to understand
- Keep the benefits information up to date

What are challenges we faced?

- Time constraints- Quickly design a training environment for developer apprentices (Code the dream’s mission) and functional MVP for funding proposals.
- No user interviews were done before designs

What was my role and who was part of the team?

I was the sole UX designer/researcher collaborating with the stakeholders and a team of 10 developers, project managers, and a senior public benefits analyst.

What was the process you followed?

The design process began with defining the problem- which was designing the AI chat and admin screens for the developer team. I then did a competitive analysis for the current AI chatbots and then received a list of user matrix roles and some existing wireframes for what the developers needed for document ingestion/AI testing.

This helped me understand the process they need to go through to make the AI chat work. With that information, I went on to create 4 userflows.

Next I created some sketches based off my competitive analysis - ChatGPT, ParentalLeave, NotebookLM.

These concepts were further developed into low- and high-fidelity prototypes using Figma and Miro. Finally, in the testing and iteration phase, I conducted internal testing and refined the designs based on the feedback received. I also prepared and carried out a user research study involving five benefits navigators and found their painpoints and wants within the high fidelities

What was the outcome?

With that being said, we’re continuing to refine the platform based on navigator and PBIF feedback, and then in the future, moving into statewide scaling, while supporting national replication.