Virtual Interviewer

Overview

Virtual AI is full stack platform that simulates realistic job interviews using real time voice, text, video, and code evaluation. Powered by an intelligent LangGraph ReAct agent with Google Gemini, Virtual AI dynamically adapts its questions based on your uploaded resume, selected job level, and interview mode.Also candidates receive immediate feedback, spoken questions via speech synthesis, and post interview analytics.

Table of Contents

• Features
• System Architecture
• Tech Stack
• Database Schema
• Project Setup
  • Backend Setup
  • Frontend Setup
• Live Links
• Team Members & Mentors
• Screenshots Showcase

Features

• Real Time Voice & Speech Interaction: Hands free speech-to-text (STT) transcription and automated speech synthesis (TTS) for natural interviews.
• LangGraph ReAct AI Agent: Context aware AI interviewer that dynamically formulates targeted follow up questions, evaluates technical accuracy, and scores candidate turns in real time.
• Zero Downtime LLM Failover: Primary LLM integration with Google Gemini, automatically failing over to Llama-3.3-70B without breaking active sessions.
• Resume PDF Parsing & Context Injection: Extracts resume text via pdfplumber and injects up to 3,000 characters of background context into the AI agent prompt.
• Customizable Settings & Modes: Supports Entry, Intermediate, and Senior experience levels with distinct evaluation modes.
• Post Interview Analytics: Evaluates overall, technical, communication, and confidence scores alongside word pace (WPM), filler word frequency, power words, strengths, and gives tips based on it.

System Architecture

Tech Stack

Category	Technologies
Frontend	React 18, Vite, Tailwind CSS, Lucide React, Monaco Editor, SpeechRecognition API
Backend	Python, Django, Django REST Framework
AI / NLP Agent	LangGraph ReAct Framework, LangChain, Google Gemini, Llama-3.3-70B
Auth & Storage	Supabase Auth (JWT), Supabase Storage Buckets (resumes, profile-images, videos)
Database	PostgreSQL (Supabase) & Django ORM)
PDF Processing	pdfplumber

Database Schema

Project Setup

1. Backend Setup

# Navigate to backend directory
cd backend

# Create & activate a virtual env
python -m venv venv
source venv/bin/activate  

# Install dependencies
pip install -r requirements.txt

# Run migrations
python manage.py migrate

# Start backend server
python manage.py runserver

2. Frontend Setup

# Navigate to frontend directory
cd frontend

# Install dependencies
npm install

# Start the server
npm run dev

Live Links

• Hosted Website: usevirtual-ai.vercel.app
• Hosted Backend: virtual-ai-iimu.onrender.com
• Demo Video: Demo Video

Team Members & Mentors

Team Members

Name	Email	GitHub
Om Mukherjee	om17092004@gmail.com	@meekhumor
Aryan Yadav	aryanyadavgr10@gmail.com	@Aryan-y-77
Aman Vatsa	amanvatsa13@gmail.com	@amanv13

Mentors

Name	Email	GitHub
Harshala Mahajan	mharshala334@gmail.com	@harshala334
Sayali Khandare	sayalisayali924@gmail.com	@Toppersayali

Screenshots

Landing Page

Interview Settings

Dashboard

Upload Resume

Review

Feedback

Analysis

Courses

Skill based Interview

Job based Interview

Interview Simulator