🤓 Latest Submissions
Interview Lab
Interview Lab is a voice-based mock interview coach that helps job seekers practice out loud instead of just reading questions off a screen. Most interview prep tools are text-based — you type an answer, read feedback, repeat. But real interviews happen through speech: hesitation, pacing, and how you think on your feet all matter, and none of that shows up in text. We built this using AssemblyAI's Voice Agent API as the real-time conversational engine. A single WebSocket connection handles speech-to-text, LLM reasoning, and text-to-speech together, along with turn detection and interruption handling — so the interviewer can ask a question, listen to your spoken answer, and respond with a natural follow-up, just like a real conversation with a human interviewer. On top of that real-time voice loop, we use OpenAI to handle the parts that don't need to happen live: generating a relevant set of interview questions based on the role you're preparing for, and analyzing the full session transcript afterward to produce a written feedback report — covering answer structure, clarity, filler words, and concrete suggestions for improvement. The main use case is job interview practice (behavioral and technical rounds). Our goal is to make interview prep feel less like a quiz and more like a rehearsal — so by the time you're in the real interview, the format itself isn't the thing you're nervous about.
30 Sep 2026
