Is Databricks intended to be used in customer facing application architectures?  I have heard that Databricks is primarily intended to be internally facing.  Is this true?

If you are using it for customer facing ML applications, what tool stack are you using?