is there a way to properly clean the email before RAG, i find signature poisoning the RAG because those are in the body of the email and so far no easy way to clean them
any of you solved this?
Hello, I am not sure I think if you are getting this through a traditional email node you could use some kind of Formatting node or Regex to get rid of the signature if you find a pattern. I also think that parsing that part of the HTML is a good solution as well. Then you can use that output inside of the RAG agent. I hope this helps.