Providing Context
So how did we get here?
Capstone Project
As a senior in Informatics, I chose to partner with a real company for my 6-month Capstone project to gain meaningful, real-world experience and take on a more challenging design problem.Who did I work with?
Collaborated with 4 teammates: 1 researcher, 1 PM, and 2 engineers
Sponsored by a Principal Engineer, Data Science Manager, and UX DesignerZ.AI is…
an internal tool that allows users to quickly experiment with prompt engineering using secure Customer Support Data and Azure OpenAIZ.AI Empowers users to safely and efficiently provide their customer with world class supportZ.AI is…
also challenging to learn and difficult to use. It’s usability challenges and steep learning curve negatively impact early adoption of the tool. These limitations constrain the tools scalability and in turn limit how well Microsoft can better support their customersProviding Context
Why does Z.AI matter?
Enhance Support Efficiency
Streamline workflows and reduce errors to improve troubleshooting and issue resolution.Empower Support Teams
Provide better AI-driven tools for faster, more accurate decision-making.Drive AI Innovation
Advance Microsoft’s leadership in AI-powered customer service.
Research to solutions process
Who we talked to
We identified two key user personas through surveys and interviews, each with distinct needs, pain points, and motivations for using the platform.
Research Methods
We conducted user surveying (30 responses), User testing of the current product (10 users), and a UX heuristic evaluation to understand key points that were preventing effective use of Z.AI
Opportunities for improvement
Based on these findings, we identified four clear opportunities for improvement that would transform the user experience.
User Flows
I then worked with my PM to develop the user flow for each persona's use case (Consumer and Creator)Use Case #1
Use Case #2
Sketches
Following up on my sketches and initial research I designed some low fidelity mockups to help drive our design direction and gain an initial idea of what we want the new product to look likeLo-Fidelity Wireframes
Proposed Solution
Home Page Improvements
We simplified navigation by centering core actions, adding a search bar, and reducing the sidebar to five tabs. A shorter intro and new Social Hub make the experience clearer and more engaging, addressing clutter and overload.Before
After
I simplified the AI experiment creation process by reducing steps from eight to five and adding visual progress tracking. Clear guidance and tutorial videos now support users throughout, with a new tutorial button for quick help. These changes make the flow faster, easier, and less overwhelming—especially for new users.Create Page Improvements
Before
After
AI Chat Page Improvements
We reduced text to improve clarity and explained key terms like temperature and top-p through tooltips. The chat box is now at the bottom for a more immersive feel, with added guidance and suggested prompts to help users use AI chat more effectively.Before
After
Q&A Page Improvements
We redesigned the Q&A page into a new Learn page with dedicated sections for tutorials, Q&A, community discussions, and office hour recordings. This centralized hub provides clearer guidance, uses more visuals and less text, and helps ease onboarding by reducing cognitive load.Before
After
Next Steps
In June our project was handed off to Microsoft for continued development and integration 
Our Impact
Our redesign will impact Z.AI’s 17,000 users and aims to grow that number post-launch. By simplifying onboarding, using plain language, and applying progressive disclosure, we’re making the platform easier to adopt and navigate. This will reduce drop-offs and improve support efficiency by enabling faster, more effective use of AI in customer workflow
Capstone Showcase
For the Informatics Capstone showcase, we set up a display with a poster, prototype walkthrough video, and a video prototype. I designed the poster layout, filmed the walkthrough, helped create the prototype video, and presented the features I led along with the ideation phase. Our design received positive feedback, and we had great conversations with judges and guests throughout the event.
