A student stands in front of a printing station, fiddling with their phone, staring at the printer, and wondering, “Why is this thing not working?” as they try to print off a document before heading to class.
What are their options?
Now, a new option: Use the QR code on the printer to connect with a customized version of the Purdue Virtual Assistant chat bot, which is pre-loaded with options to walk you through the most common printer problems: Are you at the right printer? Is your job set to the correct paper size? Are you connected to the right network? And so on...all to improve the customer experience, reduce tickets for routine issues, and solve customer issues at the point of contact.
Launched in late-February, the bot’s impact on the number of calls to the Learning Spaces labs team is still too early to tell. However, queries to the bot have averaged more than 30 conversations per week, with users seeking instructions on how to print, how to find the right printer, select the right paper size, or add funds to their print quota. Students can use the bot to report if the printer is out of paper, if it is jammed, or if the print quality is lacking.
The bot’s responses are also editable and can be adapted if data shows users asking about new issues, or if procedures change, meaning the bot can grow and change over time.
The Purdue IT Service Management team is looking for other Purdue IT teams interested in creating a custom version of the chat bot to help tackle their customers’ issues. The chat hyperlink can be shared in email signatures, on websites, inside applications, via QR code, or however best suited to attract customers’ attention and meet their needs.
If you have an idea for a customizable chatbot link, or want to learn more, contact itservicemanagement@purdue.edu or click this link here to submit a request via Team Dynamics.
The Purdue Virtual Assistant chatbot can be found on the it.purdue.edu, service.purdue.edu, rcac.purdue.edu, and purdue.edu/securepurdue websites. It uses Ivy.aI software to scan those webpages and a language learning model to help connect user questions with the available information.
Despite Ivy.AI’s advanced technology, it is only as capable as the knowledge it has available. Even when information is present, Ivy can struggle to find it because of the way it is written. The best way to improve the bot’s responses is to improve and add to the knowledge base articles it draws from, paying special attention to capture the customer’s context (and not technical jargon) of the issue.
If you have questions about the way Ivy responds to inquiries affecting your customers, contact itservicemanagement@purdue.edu or click this link here a Team Dynamics form to request a consultation – they can help develop new knowledge base articles to address missing information, show you what customers are searching after being unable to find it on your webpages, and develop strategies to improve knowledge curation.