🎙️ Prepwise – AI Mock Interview Platform

An AI-powered mock interview platform built with Next.js, Firebase, Vapi Voice Agents, and Google Gemini.

Prepwise allows users to simulate real job interviews using AI voice agents, receive structured feedback, and track their performance through a personalized dashboard.

---

🚀 Live Demo

Link:

---

🧠 Overview

Prepwise is a full-stack web application designed to help users prepare for job interviews through interactive AI voice sessions.

The platform:

• Generates interview questions dynamically
• Conducts voice-based mock interviews
• Transcribes conversations in real time
• Uses AI to generate structured performance feedback
• Stores interview history for review and improvement

This project demonstrates real-world integration of:

• AI voice workflows
• Generative AI APIs
• Authentication systems
• Scalable backend architecture
• Modern UI component systems

---

🛠 Tech Stack

Frontend

• Next.js (App Router)
• Tailwind CSS
• shadcn/ui

Backend

• Next.js API Routes
• Firebase (Authentication + Firestore)
• Firebase Admin SDK

AI & Voice

• Vapi AI (Voice Agent Workflows)
• Google Gemini (Question generation + feedback analysis)

Validation

• Zod

---

✨ Features

🔐 Authentication

• Email/password authentication via Firebase
• Protected dashboard routes
• Persistent session handling

🎙 AI Mock Interviews

• Dynamic interview generation
• AI voice agent conducts the interview
• Real-time transcript capture

🧠 AI Feedback Engine

• AI analyzes transcript

• Structured feedback:

  • Communication
  • Technical clarity
  • Confidence
  • Areas of improvement

• Score breakdown system

📊 Dashboard

• View all past interviews
• Track improvement over time
• Access detailed transcripts and feedback

📱 Responsive UI

• Fully responsive layout
• Clean, modern interface
• Accessible component system

---

📂 Project Structure

prepwise/
│
├── app/
│   ├── (auth)/
│   │   ├── sign-in/
│   │   └── sign-up/
│   │
│   ├── (root)/
│   │   ├── dashboard/
│   │   ├── interview/
│   │   │   └── [id]/
│   │   │       ├── page.tsx
│   │   │       └── feedback/
│   │   │           └── page.tsx
│   │
│   ├── api/
│   │   ├── vapi/
│   │   │   └── generate/
│   │   │       └── route.ts
│   │   └── interviews/
│   │
│   ├── layout.tsx
│   └── page.tsx
│
├── components/
│   ├── ui/
│   ├── InterviewCard.tsx
│   ├── FeedbackCard.tsx
│   └── Navbar.tsx
│
├── lib/
│   ├── actions/
│   │   └── general.action.ts
│   ├── firebase/
│   │   ├── client.ts
│   │   └── admin.ts
│   ├── utils.ts
│   └── validators.ts
│
├── public/
│
├── styles/
│   └── globals.css
│
├── .env.local
├── tailwind.config.ts
├── next.config.js
└── package.json

---

⚙️ Environment Variables

Create a .env.local file in the root directory:

# Vapi
NEXT_PUBLIC_VAPI_WEB_TOKEN=
NEXT_PUBLIC_VAPI_WORKFLOW_ID=

# Google Gemini
GOOGLE_GENERATIVE_AI_API_KEY=

# App URL
NEXT_PUBLIC_BASE_URL=

# Firebase (Client)
NEXT_PUBLIC_FIREBASE_API_KEY=
NEXT_PUBLIC_FIREBASE_AUTH_DOMAIN=
NEXT_PUBLIC_FIREBASE_PROJECT_ID=
NEXT_PUBLIC_FIREBASE_STORAGE_BUCKET=
NEXT_PUBLIC_FIREBASE_MESSAGING_SENDER_ID=
NEXT_PUBLIC_FIREBASE_APP_ID=

# Firebase (Admin)
FIREBASE_PROJECT_ID=
FIREBASE_CLIENT_EMAIL=
FIREBASE_PRIVATE_KEY=

Make sure:

• Firebase Authentication is enabled
• Firestore database is created
• Vapi workflow is configured
• Gemini API key has proper access

---

🧪 How It Works

1️⃣ Interview Creation Flow

1. User selects job role
2. Backend calls Gemini API
3. Structured interview questions are generated
4. Vapi workflow is initialized
5. Voice session begins

---

2️⃣ Interview Feedback Flow

1. Transcript collected after interview

2. Transcript sent to Gemini

3. AI generates:

   • Overall score
   • Category scores
   • Improvement suggestions

4. Feedback stored in Firestore

5. Displayed on feedback page

---

🏗 Installation & Setup

1. Clone Repository

git clone https://github.com/kindajayant/ai-interview.git
cd prepwise

2. Install Dependencies

npm install

3. Add Environment Variables

Create .env.local and add your credentials.

4. Run Development Server

npm run dev

Visit:

http://localhost:3000

---

📈 Future Improvements

• Interview difficulty levels
• AI-generated resume-based questions
• Analytics dashboard with charts
• Leaderboard system
• Performance trend visualization
• Role-based interview templates
• Real-time speech sentiment analysis

---

🧱 Architectural Decisions

• App Router used for scalable routing
• Firebase Admin SDK for secure server operations
• Action-based abstraction in lib/actions
• Component-based UI architecture
• Reusable AI prompt utilities
• Zod schema validation for API inputs

---

🧩 Key Learnings

• Voice AI integration in production systems
• Secure Firebase Admin usage in Next.js
• Designing AI prompts for structured output
• Managing async voice workflows
• Building scalable full-stack architecture

---

👨‍💻 About Me

Jayant AI Engineer Intern Focused on building AI-powered systems, scalable web applications, and backend architecture.

You can connect with me here:

• LinkedIn: (add link)
• GitHub: (add link)
• Portfolio: (add link)