Agentic AI Voice Interview Platform

The Agentic AI Voice Interview Platform is a production-ready system that conducts intelligent, voice-based AI interviews by analyzing job descriptions and candidate resumes. This platform combines a sophisticated multi-agent architecture with free, open-source speech-to-text (STT) and text-to-speech (TTS) models to create a seamless voice interview experience.

Features

• Multi-Agent Orchestration: Powered by LangGraph, CrewAI, and Agno.
• Natural Voice Interaction: Uses Groq (Llama 3 70B) for intelligence and Edge-TTS for high-quality, free voice synthesis.
• Smart Knowledge Retrieval: Integration with ChromaDB for context-aware Interviewing.
• Real-time Performance: WebSocket-based communication for low-latency voice interaction.

Prerequisites

• Python 3.11 (Required for compatibility with all AI and voice libraries)
• Node.js & npm (For the React frontend)
• Groq API Key (Set in .env)
• NVIDIA API Key (For STT processing)

Setup and Execution

Please see setup.md for the detailed step-by-step setup guide.

Quick Start:

1. Environment Setup: Copy your API keys into a .env file at the root and under backend/.

2. Backend:

   python3.11 -m venv venv
   source venv/bin/activate
   pip install -r requirements.txt
   pip install reportlab
   
   # Run tests
   python test_all_features.py
   
   # Run server
   cd backend
   ../venv/bin/uvicorn app.main:app --reload --port 8000

3. Frontend:

   cd frontend
   npm install
   npm run dev

   The frontend will be available at http://localhost:5173.

Environment Configuration

Ensure your backend/.env file contains the following:

• GROQ_API_KEY
• DEEPGRAM_API_KEY
• NVIDIA_API_KEY
• DATABASE_URL=sqlite:///./interview_platform.db

Tech Stack

• Backend: FastAPI, LangChain, LangGraph, SQL Alchemy, ChromaDB.
• Frontend: React, Vite, TailwindCSS, Framer Motion.
• Voice: Edge-TTS (TTS), NVIDIA Riva/Deepgram (STT).