It would be useful for the root cause analysis plugin to be able to refer to previous root cause analysis in the database (as well as other failure). To prevent an exhaustive search over the database, each root cause analysis in the future will be run through a embedding (or reranking model) and then inserted into a pgvector table. Then in the future the root cause analysis will take the top k vectors and include their corresponding root cause analysis in the context.
It would be useful for the root cause analysis plugin to be able to refer to previous root cause analysis in the database (as well as other failure). To prevent an exhaustive search over the database, each root cause analysis in the future will be run through a embedding (or reranking model) and then inserted into a pgvector table. Then in the future the root cause analysis will take the top k vectors and include their corresponding root cause analysis in the context.