Automate Movie Recommendations with n8n, OpenAI, MongoDB

Opening Problem Statement

Meet Sarah, a film enthusiast and a small cinema club coordinator. Every week, she spends several hours sifting through movie databases and online forums to pick exciting movies perfectly matched to her club members’ interests. On top of that, she struggles to manage favorite picks from her members efficiently. Often, these manual tasks lead to missed suggestions and slow responses, frustrating the club’s vibrant community.

This struggle highlights a very specific problem: the manual process of gathering, filtering, and managing movie recommendations wastes time, introduces errors, and misses opportunities to engage movie fans. Sarah needs a smart, automated assistant to handle these tasks without exhaustive manual effort.

What This Automation Does

This n8n workflow acts as an intelligent movie recommendation assistant powered by OpenAI’s language model and MongoDB’s robust querying capabilities. When a user sends a chat message querying movie suggestions, here’s what happens:

• Receives chat input through a public webhook trigger designed for live message intake.
• Processes the chat message with OpenAI’s Chat Model to understand the user’s intent.
• Uses OpenAI to dynamically generate a precise MongoDB Aggregation Pipeline query tailored to the user’s request.
• Executes the aggregation pipeline against a MongoDB movies collection to fetch relevant movie data, such as highly rated films or films matching specific genres or criteria.
• Maintains a contextual window memory of recent interactions for better conversational continuity.
• Allows users to save their favorite movie picks directly back into MongoDB through an automated insert workflow.

Thanks to this workflow, Sarah and her club members can save hours previously spent on manual searches, eliminate errors in data retrieval, and foster higher engagement by quickly responding to movie queries with intelligent, personalized recommendations.

Prerequisites ⚙️

• An active n8n account with webhook capabilities.
• OpenAI API Key to enable the Chat Model node for natural language processing.
• A MongoDB database configured with a movies collection containing fields such as plot, genres, cast, title, rating, and more.
• MongoDB credentials securely stored in n8n to allow query execution and inserts.
• Basic familiarity with n8n workflows and node configurations is helpful but not required.

Step-by-Step Guide to Build This Automation

1. Set Up the Chat Message Webhook Trigger

Navigate to Nodes > Add Node > Langchain > Chat Trigger. Configure it in Webhook mode with public access allowed from all origins (*). This node listens for incoming chat requests and triggers the workflow.

After saving, note the webhook URL to test sending chat messages later. You should see the webhook active with a unique ID.

Common Mistake: Forgetting to set Allowed Origins can block external chat clients.

2. Configure the OpenAI Chat Model Node

Add the node named OpenAI Chat Model from Langchain > LM Chat OpenAI. Link your OpenAI API credentials, which n8n will use to process natural language inputs.

This node interprets user chat messages and helps decide what query or data retrieval is needed.

Common Mistake: Entering incorrect API keys or missing credential assignment will cause authentication errors.

3. Apply Window Buffer Memory for Context

The Window Buffer Memory node from Langchain adds context awareness to conversations by keeping the last 10 exchanges visible. This is essential for more natural, continuous dialog.

Set contextWindowLength to 10. This means the AI remembers the last 10 chat inputs when generating responses.

Common Mistake: Setting the context window too small might cause disjointed replies.

4. Create the MongoDB Aggregate Node to Fetch Movies

Select MongoDBAggregate from n8n-nodes-base.mongoDbTool. This node runs aggregation queries on your MongoDB collection.

Configure it to use your MongoDB credentials, set the operation to aggregate, and specify your movies collection.

The key clever part: the query parameter is dynamically set by the AI, using a special expression {{ $fromAI("pipeline", "The MongoDB pipeline to execute" , "string" , [{"$match" : { "rating" : 5 } }])}}. This instructs the AI to generate exactly the right aggregation pipeline JSON based on chat input.

Common Mistake: Incorrect collection names or malformed aggregation pipelines will cause the node to fail.

5. Integrate the AI Agent for Movie Recommendations

Use the AI Agent – Movie Recommendation node from Langchain agent category. This powerful agent orchestrates the chat input, selects appropriate tools, and directs responses.

Configure its prompt with instructions to use the MongoDBAggregate node for querying, provide the user’s chat input, and insert favorites only when confirmed, formatted as { "title" : "