ValidAIte: GenAI Testing Platform
End-to-end UX for ValidAIte, a Generative AI testing platform that folds fragmented, manual QA work into a single, dashboard-driven product.
The Problem
QA engineers were working across a patchwork. Different tools, different tabs, nothing that talked to each other. Testing knowledge lived in fragments, manual, hard to trace, easy to lose, and every handoff between them added friction and a quiet cognitive tax. The work wasn't just slow. It was scattered.
The Research
So I started with the people doing it. I sat with QA engineers for interviews and task analysis, watched where the real friction lived, and wrote up findings that fed straight into the information architecture and the roadmap. Four engineers, a small sample, but the patterns were hard to miss.
The Design
From there, I owned the design. I defined the IA, made the design decisions, and built high-fidelity Figma prototypes for one unified, dashboard-driven ecosystem, with senior designers reviewing and signing off along the way. Underneath it all sat a 50+ component modular design system, built as a single source of truth so the product could grow without drifting.
The Outcome
And it shipped. The patchwork of disconnected tools became one dashboard-driven workflow that went live, fragmented, manual QA finally pulled into a single place to work. The design system gave the team one language to build in, and the research gave them a roadmap grounded in how engineers actually test, not how we'd assumed they did.
Rakunagare