Customer Experience & Support
A layered support model of chatbots, fuzzy search, and a digital assistant that reduced support tickets by 70% year over year.
The Challenge
Customers reached support in many ways, not just through one website. Inquiries spanned multiple sites and product lines and arrived through several channels. They ranged from general questions to account-specific requests, technical issues, and repetitive tasks, and no single tool could cover all of it.
The Approach: A Layered Support Model
I designed three components, each handling a different part of the customer journey:
- Website chatbots (Microsoft Copilot Studio agents). Answer common questions on the websites so customers can resolve issues without contacting support.
- Fuzzy search. Helps customers find the right information even when their wording doesn't match the content.
- Digital assistant. Handles most of the requests that need customer-specific context by drawing on customer and support history. It also processes inquiries that arrive from outside the websites: identifying spam and duplicates, merging follow-up tickets, routing requests to the right teams, and answering predictable questions.
The result
70% year-over-year reduction in customer and technical support tickets.
Chatbots and search resolve questions before a ticket is ever opened, and the digital assistant handles the ones that come through, so support staff can spend their time on issues that need a personal touch.
Have a similar problem? Let's talk →