Better Service Management with
Artificial Intelligence
Arvind Ganga
a.ganga@topdesk.com
Contents
AI: what are we talking about?
AI in Service Management
TOPdesk and AI
AI is not new
1950s: Robot arm 1960s: first chatbot
1980s: computers generating art 1990s: computer beats chess masters
Machine Learning
• Current technology that drives AI
• Computer learns from examples
• Requires sufficient examples
• Computer does not actually become
intelligent, but effectively learns how
to perform a narrowly defined
specific task
For example: to distintuish
dogs from muffins, give
many examples of dogs
and muffins
Or parrots and guacamole Puppies and fried chicken
Or sheep dogs and mops
Machine Learning never stops learning. Every time you use it and give it feedback, it will improve.
Three main applications of machine
learning
• Classification / categorization
• Natural Language Processing
• Recognizing trends, and recognizing
whether a pattern fits a trend
AI is already all around us. We all use it on a daily basis, whether we realize it or not
Examples of common AI
• Recommendation systems give you
recommendations based on your personal
preference
• Spam filtering
• Facebook brings you news based on what
you, and your friends, like
• Google Maps not only routes you
efficiently to your destination, it also gives
notifications about traffic jams, road
blocks, etc
AI becoming standard
• AI helps people be more effective
• People use AI in their daily lives
• They experience the value, and
get used to it
• They will start expecting AI in their
software at work
AI helps people be more efficient
• AI takes repetitive tasks out of
the hands of humans
• It does these small tasks
better than humans
• It supports humans by giving
suggestions and notifications
• Humans get more effective by
using AI
• Because of AI, humans can
focus on creative and social
aspects of their work
What have we learned?
AI is about machine learning
AI is about helping people
AI relies on sufficient examples
Users expect AI in their office software
Contents
AI: what are we talking about?
AI in Service Management
TOPdesk and AI
Three roles in Service Desk
End user Operator Manager
End user
AI helps end user to help herself:
• Unlock knowledge base because AI
better understands user questions
• Offer help 24-7
Operator
• AI takes over difficult and
error-prone registration:
categorization, routing, impact
analysis, linking SLA
• AI selects most important call
to work on using predictions
based on historical knowledge
and language analysis
• AI facilitates solving calls by
suggesting solutions
Manager
• AI alerts on anomalies:
unexpected call volume, many
calls about one topic, or from
one department/branch
• AI suggests planning based on
predicted and actual call
volumes and categories
• AI identifies training
opportunities to help your
employees to grow further
Next Level
Service Management
AI lets you bring Service Management
to the next level:
• End users can help themselves better
and around the clock
• Registration, selecting calls, and
solving calls becomes easier for
operators
• Managers are alerted on anomalies,
and have skilled people working
where they are needed most
Contents
AI: what are we talking about?
AI in Service Management
TOPdesk and AI
Agile introduction of AI in TOPdesk
• AI has many unknowns: technically, functionally, adoption by
users, …
• TOPdesk introduces AI in small steps, in different modules, SSP
or secure, Labs, Marketplace
• Listen to feedback and proceed accordingly
Our first AI solution categorizes incidents based on request and brief description. This
comes available to everyone soon
Our Worcade platform just introduced a chat bot that
gives an interactive tutorial to new users. This is a
first step in using chat bots.
To facilitate prioritization, we are investigating how to use the data in the audit trail to predict which incidents won’t
meet the required resolution time, or switch operator repetitively.
We are investigating whether AI can detect sentiment/emotion at customer contact, and use that for prioritization
Chat bots are a hot topic, but the exact use case differs per customer. We are
investigating if and how we can deploy chat bots as a standard solution
Current solutions and ideas are based on data of single customer. This has
drawbacks: new customers may not have sufficient data to get started, or there may
not be enough data to learn all tasks
SaaS customers can benefit from each other’s knowledge and experience. We will
investigate how to use this data effectively. Our first priority is the privacy and
security of our customers’ data.
What will TOPdesk do?
Already building solutions
Researching new functionality
In product, labs, and marketplace
Expect AI features in the coming year!
#ToT18#ToT18
Thank you!
https://www.linkedin.com/in/arvindganga/
Arvind Ganga
a.ganga@topdesk.com
Better Service Management with AI

Better Service Management with AI

  • 1.
    Better Service Managementwith Artificial Intelligence Arvind Ganga [email protected]
  • 2.
    Contents AI: what arewe talking about? AI in Service Management TOPdesk and AI
  • 3.
    AI is notnew 1950s: Robot arm 1960s: first chatbot 1980s: computers generating art 1990s: computer beats chess masters
  • 4.
    Machine Learning • Currenttechnology that drives AI • Computer learns from examples • Requires sufficient examples • Computer does not actually become intelligent, but effectively learns how to perform a narrowly defined specific task
  • 5.
    For example: todistintuish dogs from muffins, give many examples of dogs and muffins
  • 6.
    Or parrots andguacamole Puppies and fried chicken
  • 7.
    Or sheep dogsand mops
  • 8.
    Machine Learning neverstops learning. Every time you use it and give it feedback, it will improve.
  • 9.
    Three main applicationsof machine learning • Classification / categorization • Natural Language Processing • Recognizing trends, and recognizing whether a pattern fits a trend
  • 10.
    AI is alreadyall around us. We all use it on a daily basis, whether we realize it or not
  • 11.
    Examples of commonAI • Recommendation systems give you recommendations based on your personal preference • Spam filtering • Facebook brings you news based on what you, and your friends, like • Google Maps not only routes you efficiently to your destination, it also gives notifications about traffic jams, road blocks, etc
  • 12.
    AI becoming standard •AI helps people be more effective • People use AI in their daily lives • They experience the value, and get used to it • They will start expecting AI in their software at work
  • 13.
    AI helps peoplebe more efficient • AI takes repetitive tasks out of the hands of humans • It does these small tasks better than humans • It supports humans by giving suggestions and notifications • Humans get more effective by using AI • Because of AI, humans can focus on creative and social aspects of their work
  • 14.
    What have welearned? AI is about machine learning AI is about helping people AI relies on sufficient examples Users expect AI in their office software
  • 15.
    Contents AI: what arewe talking about? AI in Service Management TOPdesk and AI
  • 16.
    Three roles inService Desk End user Operator Manager
  • 17.
    End user AI helpsend user to help herself: • Unlock knowledge base because AI better understands user questions • Offer help 24-7
  • 18.
    Operator • AI takesover difficult and error-prone registration: categorization, routing, impact analysis, linking SLA • AI selects most important call to work on using predictions based on historical knowledge and language analysis • AI facilitates solving calls by suggesting solutions
  • 19.
    Manager • AI alertson anomalies: unexpected call volume, many calls about one topic, or from one department/branch • AI suggests planning based on predicted and actual call volumes and categories • AI identifies training opportunities to help your employees to grow further
  • 20.
    Next Level Service Management AIlets you bring Service Management to the next level: • End users can help themselves better and around the clock • Registration, selecting calls, and solving calls becomes easier for operators • Managers are alerted on anomalies, and have skilled people working where they are needed most
  • 21.
    Contents AI: what arewe talking about? AI in Service Management TOPdesk and AI
  • 22.
    Agile introduction ofAI in TOPdesk • AI has many unknowns: technically, functionally, adoption by users, … • TOPdesk introduces AI in small steps, in different modules, SSP or secure, Labs, Marketplace • Listen to feedback and proceed accordingly
  • 23.
    Our first AIsolution categorizes incidents based on request and brief description. This comes available to everyone soon
  • 24.
    Our Worcade platformjust introduced a chat bot that gives an interactive tutorial to new users. This is a first step in using chat bots.
  • 25.
    To facilitate prioritization,we are investigating how to use the data in the audit trail to predict which incidents won’t meet the required resolution time, or switch operator repetitively.
  • 26.
    We are investigatingwhether AI can detect sentiment/emotion at customer contact, and use that for prioritization
  • 27.
    Chat bots area hot topic, but the exact use case differs per customer. We are investigating if and how we can deploy chat bots as a standard solution
  • 28.
    Current solutions andideas are based on data of single customer. This has drawbacks: new customers may not have sufficient data to get started, or there may not be enough data to learn all tasks
  • 29.
    SaaS customers canbenefit from each other’s knowledge and experience. We will investigate how to use this data effectively. Our first priority is the privacy and security of our customers’ data.
  • 30.
    What will TOPdeskdo? Already building solutions Researching new functionality In product, labs, and marketplace Expect AI features in the coming year!
  • 31.