Hi,

I believe that will only find similar records with the same text to then use that process towards automation or assisted automation. Such as...a new record is created with a specific description, similarity solutions will evaluate to find matches and then with the highest probability, point towards similar resolutions.

So I see where you're going with it, but that's not what this would do. I only think that would apply if you never train the model on anything else...but this one use case. So then you create a few default text records and possible train it on that for future records and if it equals those with a higher probability then do 'x'.

So...there's a possibility, but I think you may skew the model a bit as it would lean more towards this one use case.