Testlify is an AI-powered skills-assessment and interviewing platform used by companies to evaluate candidates. I owned the frontend of Testlify's AI interview modules — the interactive, real-time interfaces candidates use — working closely with the AI and backend teams to pin down how each mode should behave before building it.

What I built

AI interview interfaces (chat, voice, video): Built the frontend for the multi-modal AI interviews — the screens where candidates interact with an AI over text chat, live microphone, and video.
Media capture: Handled the browser APIs for real-time audio and video — camera and microphone permissions, live audio visualizations so candidates can see they're being heard, and client-side tuning to keep the AI's back-and-forth feeling conversational.
Interview state handling: Built the frontend state logic to cope with the unpredictable parts of a live AI interview — network drops, waiting on the AI to generate the next question, and moving through assessment steps without stalling the candidate.

Why it mattered

A polished frontend made these AI interviews usable at scale and cut the manual effort interviewers had to put in — on the order of 40% less administrative work per assessment. The AI interview modules were part of the platform's growth over this period, contributing to a meaningful lift in revenue.