AI Recruiter
An AI that conducts first-round screening interviews over a live voice call. Candidates open a link, an AI interviewer joins, asks questions, and produces a transcript and a scored report — no human scheduling required.
Features
AI voice interviewer
A real-time voice agent joins the call, asks questions, listens, and asks adaptive follow-ups.
Timed sessions
Each interview is time-boxed. At the limit the agent wraps up and ends the call for everyone.
Automatic transcript
Every call is transcribed with speaker labels — candidate vs interviewer.
AI scoring report
After the call the transcript is scored against your rubric — per-criterion breakdown plus a hire recommendation.
Recruiter dashboard
Create, edit, and follow up on multiple interview sessions from one place.
Per-session config
Role, questions, rubric, and duration are defined per interview and loaded by the agent at call time.
Stack
LiveKitOpen-source WebRTC infrastructure and LiveKit Agents — the call platform and the agent runtime.
LiveKit InferenceSpeech-to-text (Deepgram), text-to-speech (Cartesia), and the LLM, billed through LiveKit.
Next.js + ReactApp Router, server components, and API routes — recruiter UI and backend in one app.
Tailwind + shadcn/uiStyling and component primitives.
PrismaData layer for interview sessions, transcripts, and reports. SQLite in dev, Postgres in production.
Python agentThe interviewer runs as a LiveKit Agent — STT to LLM to TTS with semantic turn detection.
Compliance
AI interviewing of job candidates is regulated — e.g. NYC Local Law 144 and the EU AI Act classify it as high-risk. Disclose AI use and obtain candidate consent before any real use.