Hello! I’m using a vector database to store texts containing variables. I need to ensure that when a bot queries the database, the text is returned exactly as stored, regardless of how the query is phrased. In other words, the request may be formulated differently, but the most relevant response should be retrieved based on the overall context while keeping the text unchanged.
How can I achieve this in n8n? Which nodes should I use for semantic search and extracting the original text? I’d appreciate any advice or examples!