An AI-powered interview assistant application that helps solve coding problems using Ollama models.

• Node.js (v14 or higher)
• npm or yarn
• Ollama installed and running

brew install ollama

curl -fsSL https://ollama.com/install.sh | sh

Download the installer from ollama.com

After installing Ollama, you need to pull the required models:

ollama pull deepseek-coder:6.7b
ollama pull llava:7b

Verify that the models are installed:

ollama list

Install dependencies for both the server and the application:

# Install all dependencies at once
npm run install:all

# Or install manually:
# Install server dependencies
cd server
npm install

# Install application dependencies
cd ../app
npm install

In a separate terminal, start the Ollama server:

ollama serve

The Ollama server will run on http://localhost:11434 by default.

Verify that Ollama is running:

curl http://localhost:11434/api/tags

Use the provided startup script to launch everything automatically:

./start.sh

Or using npm:

npm start

This script will:

1. Check if Ollama is installed and running (start it if needed)
2. Verify required models are installed (pull them if missing)
3. Start the Node.js server
4. Start the Electron application

If you prefer to start services manually:

Open a terminal and navigate to the server directory:

cd server
npm start

The server will start on port 3000 (or the port specified in the PORT environment variable).

Open another terminal and navigate to the app directory:

cd app
npm start

This will launch the Electron application.

./start.sh
# or
npm start

1. Make sure Ollama is running (ollama serve)
2. Start the Node.js server (cd server && npm start)
3. Start the application (cd app && npm start)
4. Click "Start Session" in the application
5. Send LeetCode problems or coding questions

interviews/
├── app/          # Electron application (frontend)
├── server/       # Node.js server (backend)
├── start.sh      # Startup script for all services
├── stop.sh       # Stop script for all services
├── package.json  # Root package.json with npm scripts
└── README.md     # This file

To stop all running services, use the provided stop script:

./stop.sh

Or using npm:

npm run stop

This script will:

• Stop all Ollama processes
• Stop the Node.js server
• Stop all Electron application processes
• Clean up stale lock files

Or stop them individually:

• Ollama: pkill -f 'ollama serve' or press Ctrl+C in the terminal where it's running
• Server: pkill -f 'node.*server.js' or press Ctrl+C in the server terminal
• App: Close the Electron application window

If processes don't stop normally, force kill them:

pkill -9 -f 'ollama serve'
pkill -9 -f 'node.*server.js'
pkill -9 -f 'electron.*app'
pkill -9 -f 'electron-forge'

• Make sure ollama serve is running
• Check that port 11434 is not occupied
• Verify Ollama installation: ollama --version

• Pull the required models:

  ollama pull deepseek-coder:6.7b
  ollama pull llava:7b

• Check available models: ollama list

• Ensure the server is running on port 3000
• Check that no other application is using port 3000
• Verify server logs for error messages

• Make sure all dependencies are installed (npm install in both app and server directories)
• Check Node.js version: node --version (should be v14 or higher)
• Review application logs for specific errors

If you see an error like Failed to open LevelDB database: File currently in use:

This usually means another instance of the application is already running or a previous instance didn't close properly. The startup script (start.sh) automatically handles this, but if you're starting manually:

macOS:

# Kill any existing Electron processes
pkill -f "electron.*app"

# Remove stale lock file (if needed)
rm -f ~/Library/Application\ Support/app/IndexedDB/file__0.indexeddb.leveldb/LOCK

Linux:

pkill -f "electron.*app"
rm -f ~/.config/app/IndexedDB/file__0.indexeddb.leveldb/LOCK

Windows:

taskkill /F /IM electron.exe
# Then manually delete: %APPDATA%\app\IndexedDB\file__0.indexeddb.leveldb\LOCK

The application uses the following Ollama models:

• deepseek-coder:6.7b - For code generation and problem solving
• llava:7b - For vision/image analysis tasks

You can use other models by modifying the configuration in the server code.

GPL-3.0