I remember vaguely that once after the chat conversation I have received a follow up from an agent and then it was added to my tickets. So it might happen once your request is processed properly.
Thatâs logical, they wonât let some famously unreliable AI hand out support tickets⌠Thatâs for a (trained) human to do.
The AI is just to placate the customer and go through those âDid you try to switch it off and back on again?â routines: First level support people have checklists to follow (rigidly, no matter what the customerâs request actually is), and that part can indeed easily be outsourced to an AI (or even a simple MCQ).
After this initial filtering they can bother somebody who actually has some know-how and can get the case in the right pipeline, resulting in a support ticket being opened.
Pretty sure a ticket (number) is assigend automatically by the system used (in this case Zendesk). If and how ChatGPT could be integrated might be another story however there is no need for a human being to create ticket (numbers)âŚ
While the AI initial screening makes sense, the case I referred to is a few months old. So I have spoken with a human agent via chat and afterwards got a follow up and a ticket. Just commenting to set the record straight.
it looks like the ticket was created an hour after the chat got finished.
There is indeed no need for a human to create numbers. The actual need for a human would be to do triage, and prevent the cases of user error and false alerts from overloading the support system.
Itâs not like they have hundreds of support engineers twiddling their thumbs eagerly awaiting a chance to show their mettle, so somebody has to make sure they only spend their time on those cases which really require their know-how.
I am coming back here to this topic to discuss AI filtering and support improvement
So naturally it is linked to the following topics
Could we possibly imagine automatic, fast-tracking sending a phone for repair and generating a RMA? With possibly some waivers like customers accepting the loss of data and paying out of pocket if the warranty proves to be unapplicable?
In an ideal world I would not have suggested that, as it is not a proper way of handling stuff. But the situation is far from ideal and I can only imagine frustration of having a bricked/unusable phone in a drawer for months (sic!).
Like for example recently there has been an issue with fingerprint scanners and phones not booting properly. It was possible to create a special dedicated troubleshooting path (FP5 bricked after the update or alike) but from what I hear people have been still waiting weeks for RMA. Is Cordon also swamped?
Curious to hear your thoughts
I am tagging @anon76468197 but it is a more general question