Build Agentic AI Applications with Node.js and LangChain

A modern web application demonstrating how to build AI-powered applications using Node.js and LangChain. This project showcases examples for integrating AI capabilities into web applications.

Features

• 🤖 AI-powered interactions using OpenAI and LangChain
• 🚀 Modern Node.js application architecture
• 📊 PostgreSQL database integration
• 🎨 Clean and responsive UI

Prerequisites

• Node.js (v20 or higher)
• pnpm package manager
• Heroku Managed Inference and Agents Add-on: Claude-3-7-Sonnet and Cohere-Embed-Multilingual (for RAG example)
• LangSmith API key (optional for tracing)
• PostgreSQL with pgvector support (for RAG example)
• OpenWeather API key (for Agent example)

Installation

1. Install pnpm (if not already installed):

   corepack install pnpm

2. Clone the repository:

   git clone https://github.com/heroku-examples/build-ai-apps-talk
   cd build-ai-apps-talk

3. Install dependencies:

   pnpm install

4. Set up environment variables:

   • Copy .env.sample to .env
   • Fill in your API keys and configuration

Development

Start the development server:

pnpm dev

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

Deployment

Option 1: Deploy to Heroku with One Click

Option 2: Manual Heroku Deployment

Note

Make sure you have the AI plugin by running: heroku plugins:install @heroku/plugin-ai

1. Create a new Heroku app:

   heroku create <app-name>

2. Add PostgreSQL addon:

   heroku addons:create heroku-postgresql:essential-0

3. Add Managed Inference and Agents Add-on:

   heroku ai:models:create claude-3-7-sonnet --as INFERENCE
   heroku ai:models:create cohere-embed-multilingual --as EMBEDDING

Note

You can find the list of available models by running: heroku ai:models:list

4. Set up the database schema:

   heroku pg:psql -f data/schema.sql

5. Configure environment variables:

   heroku config:set OPENWEATHER_API_KEY=your_key
   ...

6. Deploy to Heroku:

   git push heroku main

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.