AI Interview Coach

AI Interview Coach is a full-stack web application designed to help job seekers prepare for interviews. It leverages the power of Google's Gemini LLM to act as an intelligent AI interviewer, providing in-depth, real-time feedback.

This project is built with a modern architecture using React/TypeScript on the frontend and Node.js/Express on the backend, containerized with Docker, and designed for deployment on the Google Cloud Platform.

The Problem This Solves

Job interviews are a high-stakes, high-stress process. Many talented candidates struggle not because of a lack of skills, but due to a lack of practice in a realistic setting. Key challenges for job seekers include:

• Lack of Realistic Practice: Practicing in front of a mirror or with friends doesn't simulate the pressure of a real interview.
• No Immediate, Objective Feedback: It's hard to know what you're doing wrong. Friends may be too polite, and professional coaches can be expensive.
• Context is King: Generic interview questions are helpful, but real interviews focus on the specific role and the candidate's actual resume.
• Accessibility: High-quality interview coaching is often a paid service, creating a barrier for many.

This AI Interview Coach directly addresses these problems by providing a free, accessible, and highly personalized platform for users to practice in a simulated, yet realistic, verbal interview environment. It acts as a dedicated coach that's available 24/7.

Key Features

• Dynamic Interview Sessions: Users can set up customized interview sessions by selecting their industry, job role, and preferred language.
• Deep Personalization: The AI uses context from an uploaded resume (PDF), a profile summary, and the job description to ask highly relevant, tailored questions.
• Structured Interview Flow: The interview progresses through distinct phases: Introduction, Experience, Role-Specific, and Personality, with a user-configurable number of questions for each phase.
• Dual AI Personas:
  • AI Interviewer: Acts as the conversationalist, asking questions naturally.
  • AI Analyst: Works behind the scenes to provide objective analysis, scores, hints, and example answers.
• Verbal Interviews (TTS & STT):
  • Text-to-Speech: The AI Interviewer "speaks" using high-quality, selectable voices from Google Cloud TTS.
  • Speech-to-Text: Users can answer questions using their voice, with real-time transcription powered by a WebSocket connection to the Google Cloud STT streaming API.
• Modern & Responsive UI:
  • A professional interface with fully functional Light & Dark modes.
  • Clean and intuitive layouts for both the setup and the interview session.
  • Smooth animations (Framer Motion) and audio feedback (sound blips) for a polished user experience.
• Comprehensive Analysis: After the interview, users receive a final summary report detailing their overall score, key strengths, and areas for improvement.

Tech Stack

Frontend

• Framework: React.js with TypeScript
• Styling: Material-UI (MUI) with custom theming
• State Management: React Hooks (useState, useEffect, useContext, Custom Hooks)
• Animation: Framer Motion
• Notifications: React Hot Toast
• Real-time Communication: Client-side WebSocket API

Backend

• Framework: Node.js with Express.js
• Language: TypeScript
• Real-time Communication: ws (Node.js WebSocket library)
• AI & Cloud Services:
  • Google Gemini API: For core conversational logic and analysis.
  • Google Cloud Text-to-Speech: To generate the AI's voice.
  • Google Cloud Speech-to-Text: For real-time streaming transcription of the user's voice.
• Utilities: multer for file uploads, pdf-parse for resume reading.

Environment

• Containerization: Docker & Docker Compose
• Deployment Target: Google Cloud Run

Local Setup & Installation

Prerequisites

• Node.js (v18 or later)
• Docker & Docker Compose
• Google Cloud SDK (gcloud CLI) installed and authenticated (gcloud auth login).
• A Google Cloud account with an active billing account.

Step 1: Google Cloud Project Setup

1. Create a Project: Go to the Google Cloud Console and create a new project. Note your Project ID.

2. Enable APIs: In your new project, you must enable the following APIs. You can find them by searching their names in the console's search bar.

   • Generative Language API (for Gemini)
   • Cloud Text-to-Speech API
   • Cloud Speech-to-Text API

Step 2: Get API Keys & Credentials

This application requires two types of credentials.

A. Gemini API Key

1. Go to Google AI Studio.
2. Click "Get API key" and then "Create API key in new project".
3. Select the Google Cloud Project you just created.
4. Your API key will be generated. Copy this key.
5. You will use this for the GEMINI_API_KEY variable in your .env file.

B. Google Cloud Service Account (for TTS & STT)

The Text-to-Speech and Speech-to-Text libraries use a service account file for authentication.

1. In the Google Cloud Console, navigate to IAM & Admin > Service Accounts.
2. Click + CREATE SERVICE ACCOUNT.
   • Give it a name (e.g., ai-coach-service-account).
   • Click CREATE AND CONTINUE.
3. Grant Permissions: Add the following roles to the service account:
   • Cloud Text-to-Speech User
   • Cloud Speech-to-Text User
   • Click CONTINUE, then DONE.
4. Generate a JSON Key:
   • Find the service account you just created in the list and click on its email address.
   • Go to the KEYS tab.
   • Click ADD KEY > Create new key.
   • Select JSON as the key type and click CREATE. A JSON file will download automatically.
5. Prepare the Credential:
   • Open the downloaded JSON file.
   • Copy its entire content.
   • You will use this for the GCP_CREDENTIALS_JSON variable in your .env file. It must be a single-line string (remove all line breaks).

Step 3: Local Configuration

1. Clone the Repository:

   git clone https://github.com/ru4ls/ai-interview-coach.git
   cd ai-interview-coach

2. Create .env file:

   • In the root directory of the project, create a file named .env.
   • Add the credentials you acquired in the previous step:

   # From Google AI Studio
   GEMINI_API_KEY=your_gemini_api_key_here
   
   # The single-line JSON content from your Service Account file
   GCP_CREDENTIALS_JSON={"type": "service_account", "project_id": "...", ...}

   • Copy this .env to /backend/.env

Step 4: Run the Application

1. Build and Run with Docker Compose: This command will build the Docker images and start both containers.

   docker-compose up --build

2. Access the Application:

   • Open your browser and navigate to http://localhost:3000.
   • The backend server will be running on http://localhost:8080.

License

Licensed under the CC BY-NC-SA 4.0 license. You can find the full license text in the LICENSE file.