Career Conversation - TypeScript Version

This is a TypeScript port of the Python career conversation chatbot. It provides the same functionality as the original Python version but runs on Node.js with Express.js instead of Gradio.

Learning Project: This project was created as part of learning from the Udemy course by Ed Donner. You can find the course here.

Features

• AI-powered career conversation using OpenAI's GPT-4o-mini
• Tool integration for recording user details and unknown questions
• Push notifications via Pushover
• PDF parsing for LinkedIn profile data
• Web-based chat interface

Prerequisites

• Node.js (version 16 or higher)
• npm or yarn package manager
• OpenAI API key
• Pushover account and API keys (optional)

Setup

1. Install dependencies:

   npm install

2. Set up environment variables: Create a .env file in this directory with:

   OPENAI_API_KEY=your_openai_api_key_here
   PUSHOVER_USER=your_pushover_user_key_here
   PUSHOVER_TOKEN=your_pushover_token_here
   

3. Prepare your profile data:

   • Add your linkedin.pdf file to the me/ directory
   • Edit me/summary.txt with your career summary

4. Build the TypeScript code:

   npm run build

Running the Application

Development mode (with auto-reload):

npm run dev

Production mode:

npm start

The application will be available at http://localhost:3000

Key Differences from Python Version

1. Web Framework: Uses Express.js instead of Gradio
2. Chat Interface: Custom HTML/CSS/JavaScript interface instead of Gradio's UI
3. PDF Processing: Uses pdf-parse instead of pypdf
4. HTTP Requests: Uses axios instead of requests
5. Environment Variables: Uses dotenv package
6. Type Safety: Full TypeScript typing for better development experience
7. Clean Architecture: Separated concerns with modular TypeScript files
8. Static Assets: Organized CSS and JavaScript in separate files

File Structure

├── src/                    # TypeScript source files
│   ├── app.ts             # Main chatbot logic and Me class
│   ├── server.ts          # Express.js web server
│   ├── tools.ts           # AI tool functions and definitions
│   └── types.ts           # TypeScript type definitions
├── public/                # Static web assets
│   ├── css/
│   │   └── styles.css     # Application styles
│   ├── js/
│   │   └── chat.js        # Client-side chat functionality
│   └── index.html         # Clean HTML structure
├── me/                    # Personal profile data
│   ├── linkedin.pdf       # Your LinkedIn profile as PDF
│   ├── summary.txt        # Your career summary text
│   └── README.md          # Profile setup instructions
├── dist/                  # Compiled TypeScript output
├── package.json           # Node.js dependencies and scripts
├── tsconfig.json          # TypeScript configuration
└── project.json           # Project structure documentation

API Endpoints

• GET / - Serves the chat interface
• POST /chat - Handles chat messages
  • Request body: { message: string, history: Message[] }
  • Response: { response: string }

Tool Functions

The chatbot includes two main tools:

1. record_user_details - Records when users provide contact information
2. record_unknown_question - Records questions the bot cannot answer

Both tools send push notifications via Pushover when triggered.

Development

Project Structure

The project follows a clean separation of concerns:

• Backend (src/): TypeScript files for server logic, AI integration, and tool functions
• Frontend (public/): Static assets with separated HTML, CSS, and JavaScript
• Types (src/types.ts): Centralized type definitions for better type safety
• Tools (src/tools.ts): AI tool functions and OpenAI API definitions

Making Changes

1. Backend Logic: Edit files in src/ directory
2. Styling: Update public/css/styles.css
3. Frontend Behavior: Modify public/js/chat.js
4. HTML Structure: Edit public/index.html

Building and Running

# Development with auto-reload
npm run dev

# Build TypeScript
npm run build

# Run production build
npm start

# Watch for TypeScript changes
npm run watch

To modify the chatbot behavior:

1. Edit the Me class in src/app.ts
2. Update tool functions in src/tools.ts
3. Modify the system prompt in the systemPrompt() method
4. Rebuild with npm run build

Deployment

This TypeScript version can be deployed to various platforms:

• Railway:
  • Set environment variables in Railway dashboard
  • The project includes railway.json configuration
  • Build command: npm run build
  • Start command: npm start
  • Health check endpoint: /health
• Heroku: Add a Procfile with web: node dist/src/server.js
• Vercel: May need serverless function adaptation
• DigitalOcean App Platform: Configure build and run commands
• AWS/GCP/Azure: Deploy as containerized application

Railway Deployment Notes

1. Environment Variables: Set these in Railway dashboard:

   • OPENAI_API_KEY (required)
   • PUSHOVER_USER (optional)
   • PUSHOVER_TOKEN (optional)
   • NODE_ENV=production

2. Build Process: Railway will automatically run npm install and npm run build

3. Troubleshooting Railway:

   • Check the build logs for TypeScript compilation errors
   • Verify environment variables are set correctly
   • Use /health endpoint to test if server is running
   • Ensure your me/ directory files are included in the deployment

Make sure to set environment variables in your deployment platform.

Troubleshooting

1. Module not found errors: Run npm install to install dependencies
2. TypeScript compilation errors: Check tsconfig.json configuration
3. Environment variables not loaded: Ensure .env file exists and is properly formatted
4. PDF parsing issues: Verify me/linkedin.pdf exists and is readable
5. Port already in use: Change the PORT environment variable or kill existing processes
6. Railway deployment failures:
   • Check build logs for compilation errors
   • Verify environment variables are set in Railway dashboard
   • Test health endpoint: your-app-url.railway.app/health
   • Ensure OpenAI API key is valid and has credits

License

MIT License - feel free to modify and use as needed.