AI support employee
A software company with two products wanted its support team to stop spending the day on cases the documentation already answered. Mytrya built an AI support employee that does the whole front-line job. It reads each case, finds what it needs to know, answers or asks the customer, reports bugs to engineering, hands off to the right person with a summary, follows up when the customer goes quiet, and closes the case. It works in the helpdesk, in live chat, and in a chat window on the company's own site and apps.
The problem
Most tickets could be answered from documentation that already existed. A person still had to read each one, find the article, check the account, check whether engineering already knew, write the reply and remember to follow up.
Live chat had turned into a ticket form. Almost nobody answered chats directly, so every chat became one more ticket.
What it does
- Answers from the company's own documentation, and asks or hands off when the answer isn't there
- Reports bugs to engineering and tells the customer when the fix is out
- Hands off to the right person with a written summary of what it found
- Follows up on its own when a customer goes quiet, then closes the case
- Briefs the team each morning on issues that are starting to repeat
- Drafts help articles for the questions nobody has written up yet
Where it stands
In production. On tickets a person reviews each reply before it goes out; on chat it answers customers directly. It learns from the replies the team actually sends, and a person approves each lesson before it is used. The client isn't named and its figures aren't published.