Mastra & MongoDB Demo

Multi-Tool AI Agent System

This application expands on the included Mastra weather demo to demonstrate an AI agent with multiple tool capabilities. The agent showcases Mastra's agentic, RAG, and tool capabilities through weather and movie recommendation tools.

Features

Agentic capabilities - AI agent that can handle weather and movie requests
Weather tool - Fetches current weather information for any location
Movie tool - Provides movie recommendations and information using RAG
RAG system - Semantic search across movie plots and metadata using retrieval-augmented generation
Tool integration - MongoDB Atlas Vector Search for movie embeddings
Built with Mastra framework and OpenAI

Database Structure

Movies Collection: sample_mflix.movies

Uses the MongoDB Atlas sample dataset

Embeddings Collection: sample_mflix.movie_embeddings

Vector embeddings for semantic search
Includes plot and metadata chunks

Prerequisites

Before getting started, you'll need to load the sample data in MongoDB Atlas:

Create a MongoDB Atlas cluster - If you don't have one already, create a free cluster at MongoDB Atlas
Load sample data - Load the sample datasets in your Atlas cluster:
- In your Atlas dashboard, go to your cluster
- Click "Browse Collections"
- Click "Add My Own Data" or use the "Load Sample Dataset" option
- Select and load the Sample Mflix Dataset which includes the sample_mflix.movies collection

Getting Started

Set up your environment variables in .env:

MONGODB_URI=your_mongodb_connection_string
OPENAI_API_KEY=your_openai_api_key

Install dependencies:
```
npm install
```
Create embeddings:
```
npm run create-embeddings
```
Start the development server:
```
npm run dev
```

Available Scripts

npm run create-embeddings: Create embeddings for movies in the database
npm run dev: Start the development server