Here is my entry for #wonderchallenge
I built TokenTracker
TokenTracker is a B2B SaaS concept for production AI teams. It helps them understand model spend, routing, latency, and prompt-level usage in one place, so they can see where costs are coming from and make better decisions about which model to use to reduce cost and increase efficiency.
It solves a common problem for teams shipping AI features: once they start using multiple models, it becomes hard to tell why costs are rising or which requests are driving the bill. This app brings that visibility into a single, clear product experience.
I built it as a high-fidelity multi-page design project all in Wonder. I first defined the product direction, narrowed the idea to what could be built and currently possible in Wonder, created the design system and branding foundation, and then designed each page section by section using Wonder’s tools and chat keeping brand consistency across the pages . The final result is a polished concept that feels close to production.
For the app I have built following pages —
Here is my entry for #wonderchallenge
I built TokenTracker
TokenTracker is a B2B SaaS concept for production AI teams. It helps them understand model spend,...