Success Story Manufacturing

Engineering Giant Gets 95% Accuracy in Customer Service Automation

automates Help Desk function

With over 11,000 employees and a worldwide footprint, the customer delivers factories, machinery, services, and expertise to the global cement and mineral industries. As a result, they have considerably large helpdesk operations that could greatly benefit from optimization.

The customer wanted Apexon to build a text mining model to predict the category of helpdesk tickets from the email subject line. Additionally, the customer required us to build bots for specific helpdesk divisions including IT services, application support, and self-help portals.

the Results

Key Outcomes

By using 12 resolution bots coupled with a master and trigger bot within a cognitive application, Apexon was able to dramatically improve helpdesk efficiency:

Representing 40% of the overall ticket volume
Over 500 helpdesk tickets resolved by bots, representing 40% of the overall ticket volume

90% of ticket assignments handled via SaaS RPA
90% of ticket assignments handled via SaaS RPA & touchless automation, with 95% accuracy

Headcount deployment reduced by up to 40%
Headcount deployment reduced by up to 40%

Improved CSAT scores across every service line
Improved CSAT scores across every service line

The challenge

3 key areas

Given the scale of the client’s operations and the vast number of helpdesk tickets generated daily, Apexon faced a set of very specific challenges:

Mapping & Support

Mapping & Support Mapping several potential resolution pathways and integration automation features to support them

Creating NLP-powered chatbots

Chatbots Creating NLP-powered chatbots that supported multiple languages

Successfully delivering a POC

POC Delivery Successfully delivering a POC within a 6-week timeframe

The Solution

5 key areas

By leveraging our expertise in R and Microsoft Bot Framework, Apexon was able to automate large portions of the helpdesk function, including the following tasks:

User Acknowledgment

User Acknowledgment Dispatching acknowledgment to the user on the task being processed

Problem Identification

Problem Identification Problem identification from ticket details as shared by the customer

Ticket status updates

Status Updates Ticket status updates in the IT helpdesk application (TopDesk)

Auto-resolution of each ticket

Auto-Resolution Auto-resolution of each ticket to the mandated extent

Completion Notifications

Completion Notifications Ticket and task completion notifications to the user