Will the Input and Output data of Chat with Sense go into training a general model that can be used by other customers?

Improvements to Chat with Sense won't be specific to a single customer’s Input or Output data. The Inputs and Outputs don't train any third-party model, just the Heap middleware that's interacting with OpenAI (for example, how we prompt a model with examples of phrasing that might lead it to choose to create one type of Heap chart versus another).