n8n MongoDB AI Agent for Intelligent Movie Recommendations
Overview
🎬 n8n MongoDB AI Agent for Intelligent Movie Recommendations
This n8n MongoDB AI Agent for Intelligent Movie Recommendations flow is designed to work as an AI autonomous agent that can get chat messages, query data from MongoDB using the aggregation framework. Following by augmenting the results from the sample movies collection and allowing storing my favorite movies back to the database using an “insert” flow.
🔧 Workflow Steps:
- When chat message received: Acts as the starting point, triggering the workflow upon receiving a chat message from a user.
- AI Agent – Movie Recommendation: The central AI agent that interprets user requests, manages conversation flow, and decides which tools to use for fulfilling the request, such as fetching movie data or inserting favorites.
- OpenAI Chat Model: Provides the language model capabilities for the AI agent, allowing it to understand natural language queries and generate human-like responses.
- Window Buffer Memory: Maintains a history of the conversation, enabling the AI agent to remember past interactions and provide contextually relevant responses.
- MongoDBAggregate: A tool used by the AI agent to query the MongoDB database. It uses aggregation pipelines to retrieve specific movie data based on the AI’s understanding of the user’s request.
- insertFavorite: A dedicated sub-workflow tool that allows the AI agent to add a user’s confirmed favorite movie to the MongoDB database.
📌 Use Cases:
- Intelligent movie recommendations based on criteria provided in natural language.
- Searching and retrieving movie information directly from a MongoDB database using conversational AI.
- Storing user-selected favorite movies into a database for future reference.
🧰 Required Credentials:
- OpenAI API
- MongoDB
⚙️ Notes & Enhancements:
- This workflow demonstrates the power of integrating AI agents with database operations using n8n.
- The use of MongoDB aggregation allows for complex and flexible data querying.
- The modular design with a separate ‘insertFavorite’ workflow promotes reusability and clarity.
- The workflow can be extended to include more MongoDB operations or integrate with other data sources.
Workflow Editor Screenshot
Workflow JSON Code
{
"id": "22PddLUgcjSJbT1w",
"meta": {
"instanceId": "fa7d5e2425ec76075df7100dbafffed91cc6f71f12fe92614bf78af63c54a61d",
"templateCredsSetupCompleted": true
},
"name": "MongoDB Agent",
"tags": [],
"nodes": [
{
"id": "d8c07efe-eca0-48cb-80e6-ea8117073c5f",
"name": "OpenAI Chat Model",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
1300,
560
],
"parameters": {
"options": {}
},
"credentials": {
"openAiApi": {
"id": "TreGPMKr9hrtCvVp",
"name": "OpenAi account"
}
},
"typeVersion": 1
},
{
"id": "636de178-7b68-429a-9371-41cf2a950076",
"name": "MongoDBAggregate",
"type": "n8n-nodes-base.mongoDbTool",
"position": [
1640,
540
],
"parameters": {
"query": "={{ $fromAI(\"pipeline\", \"The MongoDB pipeline to execute\" , \"string\" , [{\"$match\" : { \"rating\" : 5 } }])}}",
"operation": "aggregate",
"collection": "movies",
"descriptionType": "manual",
"toolDescription": "Get from AI the MongoDB Aggregation pipeline to get context based on the provided pipeline, the document structure of the documents is : {\n \"plot\": \"A group of bandits stage a brazen train hold-up, only to find a determined posse hot on their heels.\",\n \"genres\": [\n \"Short\",\n \"Western\"\n ],\n \"runtime\": 11,\n \"cast\": [\n \"A.C. Abadie\",\n \"Gilbert M. 'Broncho Billy' Anderson\",\n ...\n ],\n \"poster\": \"...jpg\",\n \"title\": \"The Great Train Robbery\",\n \"fullplot\": \"Among the earliest existing films in American cinema - notable as the ...\",\n \"languages\": [\n \"English\"\n ],\n \"released\": \"date\"\n },\n \"directors\": [\n \"Edwin S. Porter\"\n ],\n \"rated\": \"TV-G\",\n \"awards\": {\n \"wins\": 1,\n \"nominations\": 0,\n \"text\": \"1 win.\"\n },\n \"lastupdated\": \"2015-08-13 00:27:59.177000000\",\n \"year\": 1903,\n \"imdb\": {\n \"rating\": 7.4,"
},
"credentials": {
"mongoDb": {
"id": "8xGgiXzf2o0L4a0y",
"name": "MongoDB account"
}
},
"typeVersion": 1.1
},
{
"id": "e0f248dc-22b7-40a2-a00e-6298b51e4470",
"name": "Window Buffer Memory",
"type": "@n8n/n8n-nodes-langchain.memoryBufferWindow",
"position": [
1500,
540
],
"parameters": {
"contextWindowLength": 10
},
"typeVersion": 1.2
},
{
"id": "da27ee52-43db-4818-9844-3c0a064bf958",
"name": "When chat message received",
"type": "@n8n/n8n-nodes-langchain.chatTrigger",
"position": [
1160,
400
],
"webhookId": "0730df2d-2f90-45e0-83dc-609668260fda",
"parameters": {
"mode": "webhook",
"public": true,
"options": {
"allowedOrigins": "*"
}
},
"typeVersion": 1.1
},
{
"id": "9ad79da9-3145-44be-9026-e37b0e856f5d",
"name": "insertFavorite",
"type": "@n8n/n8n-nodes-langchain.toolWorkflow",
"position": [
1860,
520
],
"parameters": {
"name": "insertFavorites",
"workflowId": {
"__rl": true,
"mode": "list",
"value": "6QuKnOrpusQVu66Q",
"cachedResultName": "insertMongoDB"
},
"description": "=Use this tool only to add favorites with the structure of {\"title\" : \"recieved title\" }"
},
"typeVersion": 1.2
},
{
"id": "4d7713d1-d2ad-48bf-971b-b86195e161ca",
"name": "AI Agent - Movie Recommendation",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
1380,
300
],
"parameters": {
"text": "=Assistant for best movies context, you have tools to search using \"MongoDBAggregate\" and you need to provide a MongoDB aggregation pipeline code array as a \"query\" input param. User input and request: {{ $json.chatInput }}. Only when a user confirms a favorite movie use the insert favorite using the \"insertFavorite\" workflow tool of to insertFavorite as { \"title\" : \"<TITLE>\" }.",
"options": {},
"promptType": "define"
},
"typeVersion": 1.7
},
{
"id": "2eac8aed-9677-4d89-bd76-456637f5b979",
"name": "Sticky Note",
"type": "n8n-nodes-base.stickyNote",
"position": [
880,
300
],
"parameters": {
"width": 216.0875923062025,
"height": 499.89779507612025,
"content": "## AI Agent powered by OpenAI and MongoDB \n\nThis flow is designed to work as an AI autonomous agent that can get chat messages, query data from MongoDB using the aggregation framework.\n\nFollowing by augmenting the results from the sample movies collection and allowing storing my favorite movies back to the database using an \"insert\" flow. "
},
"typeVersion": 1
},
{
"id": "4d8130fe-4aed-4e09-9c1d-60fb9ac1a500",
"name": "Sticky Note1",
"type": "n8n-nodes-base.stickyNote",
"position": [
1300,
720
],
"parameters": {
"content": "## Process\n\nThe message is being processed by the \"Chat Model\" and the correct tool is used according to the message. "
},
"typeVersion": 1
}
],
"active": true,
"pinData": {},
"settings": {
"executionOrder": "v1"
},
"versionId": "879aab24-6346-435f-8fd4-3fca856ba64c",
"connections": {
"insertFavorite": {
"ai_tool": [
[
{
"node": "AI Agent - Movie Recommendation",
"type": "ai_tool",
"index": 0
}
]
]
},
"MongoDBAggregate": {
"ai_tool": [
[
{
"node": "AI Agent - Movie Recommendation",
"type": "ai_tool",
"index": 0
}
]
]
},
"OpenAI Chat Model": {
"ai_languageModel": [
[
{
"node": "AI Agent - Movie Recommendation",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Window Buffer Memory": {
"ai_memory": [
[
{
"node": "AI Agent - Movie Recommendation",
"type": "ai_memory",
"index": 0
}
]
]
},
"When chat message received": {
"main": [
[
{
"node": "AI Agent - Movie Recommendation",
"type": "main",
"index": 0
}
]
]
}
}
}