Will there be a stage where the prompt building becomes a to and fro with the AI Agent, and the co-pilot prompts the user for the information (at the start or as required), to help get all the best info? You might have called it prompting the user, but it's a bit of an overloaded term from being used in prompt engineering already.
Well, it's a start I suppose but why would anyone choose to use Einstein if they were already familiar with even the basic functionality of Tableau? If it could make a matrix of the relationship of all measures against each other and then recommend ones that had a correlation of, say, greater than .8 then that would be something useful but as is it seems a little underwhelming.
Will there be a stage where the prompt building becomes a to and fro with the AI Agent, and the co-pilot prompts the user for the information (at the start or as required), to help get all the best info? You might have called it prompting the user, but it's a bit of an overloaded term from being used in prompt engineering already.
Well, it's a start I suppose but why would anyone choose to use Einstein if they were already familiar with even the basic functionality of Tableau? If it could make a matrix of the relationship of all measures against each other and then recommend ones that had a correlation of, say, greater than .8 then that would be something useful but as is it seems a little underwhelming.