Automate Customer Support with a simple AI-Powered Ticketing System

Automate Customer Support with a simple AI-Powered Ticketing System

View Details
Link
Apps
Outlook
Outlook
OpenAI / ChatGPT
OpenAI / ChatGPT
Make
Make
Hubs
Ecommerce
Ecommerce
Case Studies
Case Studies
Unsplash Cover
Unsplash Cover
Video
General
Generate LK
Generate NL
LinkedIn Post Content
🚀 Excited to share our latest success story! We've partnered with a top UK supplier of commercial dehumidifiers to automate their customer support using an AI-powered ticketing system.

As their business grew, so did the volume of support emails. Our solution leverages AI to categorize emails based on error codes and sends tailored responses, significantly reducing the time spent on draft creation from hours to mere minutes! With 85% of emails now handled semi-automatically, our client can focus on what truly matters—providing excellent service.

Curious to learn more about this transformative project? Check out the full blog post here: Read More

#AI #CustomerSupport #Automation #Connex #DigitalTransformation #OpenAI
Hidden
Tags db
Case Study
Automation Basics
Full URL
Unsplash Cover
Unsplash Cover
KW AI GEN
AI customer support automation, email management, dehumidifiers

The Client

Top UK supplier of commercial dehumidifiers.
They wanted to automate their customer service support system.
They have a lot of emails coming through in a day asking for support their dehumidifier equipment. They wanted to use AI to categorize emails based on error codes in the email body and send customized replies based on the error categorization.

The Challenge

  • The volume of Customer Support emails was high and rising as the business grew.
  • They wanted to keep headcount low, but still there were lots of repetitive and time consuming tasks related to answer customer emails.

The Solution

  • We created workflows in Make.
  • The first scenario gets new emails and categorizes the problem as a Product problem, Order followup or others.
  • If it is a product problem the second scenario categorizes the problem using AI based on the error code. The instructions/rules to determine which error code the email falls under is specified in a document. We create an OpenAI / ChatGPT assistant based on that document.
  • Once the problem is classified the feedback email is sent to the client.

The Results

  • AI reduced the time to create drafts to emails from a couple of hours to minutes a day.
  • A human was involved to check on the answers and respond to non common inquires.
    85% of emails were handled semi automatically.

Tech Used

  • OpenAI Assistants to create drafts
Â