Hey everyone I built Round1 over the last five months, nights and weekends, alongside my day job as a product Manager. The problem started from a pattern I kept hearing about from people in staffing: agencies send candidates to clients for technical interviews, and the people deciding who gets sent — recruiters, sometimes a team lead doing a "peer review" — usually aren't developers. They're making a technical judgment call with no technical ability to make it. So agencies send 5 candidates, all 5 fail the client's L1 round, and nobody on the agency side can say why. Was it a bad candidate? A coached one? Just a mismatch? No data, just a burned relationship and burned hours. Round1 doesn't replace the recruiter or the client interview. It sits in between — after the recruiter shortlists someone, before anyone reviews them — and runs a real adaptive technical interview by voice. The candidate just clicks a link, no app, no human to schedule. The AI asks questions grounded in the actual JD, follows up on vague answers, and scales difficulty to seniority. Within about a minute of the interview ending, the recruiter gets a report: a technical score, a recommendation, and — this is the part I think is genuinely new — a separate score that flags whether the candidate's answers look coached or AI-assisted, based on response patterns and behavioral signals during the call. It's live, it's in active use with early staffing agency customers, and I'd love feedback from this community — especially if you've dealt with technical hiring, screening, or building anything in the voice-AI space. Happy to answer anything about how it works under the hood (Next.js, Supabase, GPT-4o for the interview + assessment, OpenAI TTS for voice). Try the live demo (interviewer's name is Betty) at round1.in/try — no signup needed to see it in action.

Round1

AI technical interviews that catch what recruiters can't.

AI technical interviews that catch what recruiters can't.