UXRay

Inspiration

As builders, we kept noticing the same problem. Teams often do not catch UX issues until real users run into them. Traditional user testing can take time, cost money, and be hard to access for smaller teams or fast moving projects. We wanted to build a system that could simulate realistic user behavior earlier in the process and give teams clear feedback before launch.

What it does

UXRay is an AI powered UX review system that simulates real user audiences interacting with a product. Instead of waiting for live feedback, it tests flows in advance, identifies friction points, spots confusing decisions, and surfaces weak areas in the experience. The result is actionable feedback that helps teams improve usability before those issues affect real users.

How we built it

We built UXRay as an AI driven product analysis workflow that evaluates interfaces through simulated audience behavior. The system models how different user types would move through a product, analyzes where they may hesitate or struggle, and turns those findings into practical UX recommendations. The goal was not just to generate generic critique, but to make the feedback feel specific, realistic, and useful to builders.

Challenges we ran into

One of the biggest challenges was making simulated user behavior feel believable. It was easy to generate general feedback, but much harder to create critiques that reflected how real people would actually experience a flow. Another challenge was keeping the output actionable. We wanted the system to do more than point out vague issues, so we focused on producing feedback that teams could actually use to improve their product.

Accomplishments that we’re proud of

We are proud that UXRay makes UX feedback more accessible and immediate. Instead of relying only on expensive testing or waiting for real user frustration, teams can get early insight into where their product may break down. We are also proud that the system focuses on realistic audience simulation rather than generic design commentary.

What we learned

We learned that useful UX feedback depends on context. The more grounded the simulated audience is, the more valuable the critique becomes. We also learned that product feedback is most effective when it is specific, direct, and clearly tied to the actual user experience.

What’s next for UXRay

Our next step is to make the simulations more detailed, support more product types and user flows, and improve how feedback is tailored to different audiences. Over time, we want UXRay to become a tool teams can use throughout development to test decisions earlier and build with more confidence.

*Note: *

UXRay was not kept live hosted for this submission due to backend cost constraints and hosting uptime concerns. Because the product relies on a more intensive backend workflow, we prioritized delivering a stable demo of the core experience over maintaining a continuously deployed live version.**

Built With

• browseruse
• fetch
• next.js
• playwright
• supabase