Have a look at what rooting pipeline components are available in Haystack. You can use them to route queries or documents to other nodes that can handle them best.
Distinguishing between keyword queries and natural language questions and routing them to the retrievers than can handle them best.
|RouteDocuments||RouteDocuments||Routes documents to different branches of your pipeline based on their content type or a metadata field.|
Routing table data to TableReader and text data to TransfomersReader for better handling.
Updated about 1 month ago