Turn scattered customer reviews into clear answers your team can use. It connects a vector database with AI so you can search, compare, and recommend reviews across brands in seconds. Great for marketing and product teams that need fast, reliable insights from large review sets.

A server trigger exposes five tools insert, search, compare, recommend, and list companies. A switch routes each request to the right path. For search, compare, and recommend, the flow creates embeddings with OpenAI, then calls Qdrant group search, facet listing, and recommendation APIs. Aggregate and set nodes shape clean outputs, while if nodes handle empty results with safe fallbacks. Insert and simple search use native Qdrant nodes, and a small code step turns user preferences into items for embedding.

You need a Qdrant endpoint and an OpenAI API key. The template includes steps to create the collection and a facet index for company id. Teams can cut hours of manual reading and scale to more brands without extra load. Use it for brand monitoring, competitive analysis, and quick product feedback checks.

• MCP server trigger exposes insert, search, compare, recommend, and list tools for agent clients
• Switch node routes each request by operation to the correct path
• OpenAI embeddings turn queries and preferences into vectors for better matching
• HTTP requests call Qdrant recommend, group search, and facet listing APIs
• Native Qdrant nodes handle insert and similarity search for simple operations
• Code step converts user preferences into items ready for embedding
• Aggregate and set nodes return clean, compact results to the client
• If nodes provide clear empty states when no results are found
• Manual and subworkflow triggers support testing and reuse

• Reduce manual review reading from 2 hours to 10 minutes
• Automate 80% of sorting and grouping across companies
• Improve result relevance with vector search and preferences
• Handle 10 times more reviews without extra headcount
• Connect Qdrant and OpenAI in one reliable flow

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with Qdrant and OpenAI. See the Tools Required section above for links to create accounts with these services.
3. In your Qdrant setup, note the base URL and API key. Create or choose the collection name you want to use for reviews.
4. In your OpenAI account, create an API key from the API page and keep it secure.
5. In the n8n credentials manager, create a QdrantApi credential. Enter the base URL and API key. Save the credential with a clear name.
6. In the n8n credentials manager, create an OpenAI credential. Paste the API key and save.
7. Open the List by Facet API, Group Search API, and Recommend API nodes. Select your Qdrant credential. Confirm the endpoint URL and the collection name. Update the collection name if you are not using trustpilot_reviews.
8. Open the Get Embeddings and Get Embeddings1 nodes and select your OpenAI credential. Pick the embedding model if the field is empty.
9. Run the prerequisite section. Execute Create Collection and then Create Facet Index. Check each node returns a success status.
10. Open the Qdrant MCP Server node. Set a unique path, enable authentication for security, then save and activate the workflow.
11. Test insert and search. Execute Insert Reviews with a sample review and company id. Then run Search Reviews and confirm Get Search Response returns records.
12. Validate compare and recommend. Run Group Search API and Recommend API with sample inputs. Check Simplify Group Results and Simplify Recommend Response outputs for clean data.
13. Troubleshoot common issues. If you see empty results, confirm the facet index for company id exists and that items are inserted. If you get unauthorized errors, verify the Qdrant API key and base URL. If embeddings fail, check the OpenAI credential and model name.