I built Aural because every mock interview tool I tried felt nothing like a real interview. Typing answers into a chat box might be convenient, but it does not train the part that actually matters. Thinking out loud under pressure.
So I made a voice-first mock interview platform where the AI asks follow-up questions based on your answer, then gives transcription, scoring, and feedback after the session. I also open sourced it under MIT because I wanted to stay honest about how it works, and a lot of users care about self-hosting and owning their data. The stack is Next.js, Supabase, and Postgres. Live here: https://aural-ai.com and code here: https://github.com/1146345502/aural-oss
Still early, and I am treating this as a real SaaS, not just a side project demo. My main question now is positioning. Would you frame this as a consumer interview prep product, or lean into the API and self-hostable angle for teams and developers?