Vocalis AI - Autonomous Multi-Role AI Voice Interviews for Hiring Teams & Candidates

What Vocalis AI is about

Vocalis AI is an enterprise-grade, autonomous voice interview platform powered by the official Agora Conversational AI Agent SDK (agora-agents v2.7.0), Deepgram Nova-3 ASR, Groq Qwen 3.8 27B, and MiniMax TTS over Agora SD-RTN™. Instead of a single static bot, it deploys an interactive panel of five specialized interview personas—Lead Systems Architect, Principal PM, VP of Engineering, Enterprise Customer Director, and Lead Org Psychologist—complete with sub-100ms Voice Activity Detection (VAD) barge-in floor control.

Why I built it

Most existing AI mock interview solutions are passive, single-turn text prompts that fail to emulate real engineering and leadership loops. I built Vocalis AI to bridge the gap between static prep tools and actual panel pressure. The system conducts backstage panel deliberation after every turn, actively detecting vague buzzwords, flagging resume contradictions, dynamically calibrating interview difficulty (from Foundational to Staff/Principal), and generating 360° evaluation scorecards backed by verbatim transcript quote citations.

How it helps others

• For Candidates: Delivers realistic, interruptible practice with sub-100ms voice response times, dynamic follow-up probing, interactive system design whiteboard sync, and actionable evidence-based feedback on communication and technical depth.
  Candidate Dashboard :- https://drive.google.com/drive/folders/11CB4MniZ_Jt1_x95uEcgMPSTtmG75CNb?usp=sharing

• For Recruiters & Hiring Teams: Provides an end-to-end hiring intelligence dashboard with cohort analytics, FAANG/Bar-Raiser rubrics management, and merit-plus-parity shortlisting algorithms to ensure fair, transparent, and data-backed hiring decisions.
  Recruiter Dashboard :- https://drive.google.com/drive/folders/1X8gZOp7ZSNDuBPJxe1p3AI_zBODfn0al?usp=sharing
  
  Landing Page :- https://drive.google.com/drive/folders/1X8gZOp7ZSNDuBPJxe1p3AI_zBODfn0al?usp=sharing