134
15th
Anniversary Souvenir 2019
kGw|f“} jflif{sf]T;j :dfl/sf @)&%
Shyam Krishna Khadka
Engineer
ISSD, Nepal Telecom
Introducing Artificial Intelligence into Telecom
Companies (Telco)
The telecom operators are facing newThe telecom operators are facing newThe telecom operators are facing newThe telecom operators are facing newThe telecom operators are facing new
opportunities and challenges with theopportunities and challenges with theopportunities and challenges with theopportunities and challenges with theopportunities and challenges with the
exponential growth of IoT devices and 4G/exponential growth of IoT devices and 4G/exponential growth of IoT devices and 4G/exponential growth of IoT devices and 4G/exponential growth of IoT devices and 4G/
5G network data. The challenges reside in5G network data. The challenges reside in5G network data. The challenges reside in5G network data. The challenges reside in5G network data. The challenges reside in
handling higher volumes of data and extractinghandling higher volumes of data and extractinghandling higher volumes of data and extractinghandling higher volumes of data and extractinghandling higher volumes of data and extracting
actionable insights while improving networkactionable insights while improving networkactionable insights while improving networkactionable insights while improving networkactionable insights while improving network
efficiencies and lowering operationalefficiencies and lowering operationalefficiencies and lowering operationalefficiencies and lowering operationalefficiencies and lowering operational
expenses. This is where, AI can help theexpenses. This is where, AI can help theexpenses. This is where, AI can help theexpenses. This is where, AI can help theexpenses. This is where, AI can help the
telecommunication industrytelecommunication industrytelecommunication industrytelecommunication industrytelecommunication industry
What is AI?What is AI?What is AI?What is AI?What is AI?
Artificial Intelligence (AI): Artificial +
Intelligence.
By word, AI means bringing human like
intelligence to machines, making machines think and
behave like human. The works which are tedious,
repetitive, requiring large human population can now
be simplified by using AI techniques. Artificial
Intelligence (AI) has become one of the hottest
technologies that are being used in various fields like
agriculture, industrial process, security, network,
telecommunication, health, biology etc. Some of the
examples of AI in our real life include:
1. Email Filters in Gmail1. Email Filters in Gmail1. Email Filters in Gmail1. Email Filters in Gmail1. Email Filters in Gmail
2. Chatbots2. Chatbots2. Chatbots2. Chatbots2. Chatbots
Chatbots recognize words and phrases in order
to (hopefully) deliver helpful content to customers
who have common questions. Sometimes, chatbots
are so accurate that it seems as if you’re talking to a
real person.
3. Navigation and travel apps3. Navigation and travel apps3. Navigation and travel apps3. Navigation and travel apps3. Navigation and travel apps
AI is used in apps like Google/Apple maps for
navigating, or calling an Uber, or booking a flight
ticket. Both Google and Apple along with other
navigation services use artificial intelligence to
interpret hundreds of thousands of data point that they
receive to give you real-time traffic data. When we
are calling an Uber, both the pricing and the car that
matches our ride request are decided by AI. As we
can see, AI plays a significant role in how we reach
from point A to point B.
4. Product Recommendation systems4. Product Recommendation systems4. Product Recommendation systems4. Product Recommendation systems4. Product Recommendation systems
Amazon and other online retailers use AI to gather
information about your preferences and buying
habits. Then, they personalize our shopping
experience by suggesting new products tailored to
our habits.
5.5.5.5.5. Fraud detection in financial sectorsFraud detection in financial sectorsFraud detection in financial sectorsFraud detection in financial sectorsFraud detection in financial sectors
There are many situations in financial sectors that
AI powered software/technologies are used in fraud
detection. For example: Getting SMS/Email alert if
an unusually large transaction takes place in your
account and many others.
The underlying working principle of AI lies on
data. AI needs huge amount of data, to work. AI
techniques have capacity to analyze huge data, detect
patterns among them, classify and predict them. For
a Telco like Nepal Telecom, we have large data that
may be either of customers or of network elements
(NE) data or employee data. Depending on the type
of data, AI can be used in corresponding fields.
Potential application fields of AI in telecomsPotential application fields of AI in telecomsPotential application fields of AI in telecomsPotential application fields of AI in telecomsPotential application fields of AI in telecoms
In mobile congress conference that was held at
Barcelona, Spain in 2018, "Applied AIApplied AIApplied AIApplied AIApplied AI" was one of
135
15th
Anniversary Souvenir 2019
kGw|f“} jflif{sf]T;j :dfl/sf @)&%
the eight themes of the conference. The intention was
to help the industry cut through the complexity of
artificial intelligence (AI). The same trend regarding
AI has been continued in this year of mobile congress
committee too (going to be held in February
2019). The event theme is "IntelligentIntelligentIntelligentIntelligentIntelligent
ConnectivityConnectivityConnectivityConnectivityConnectivity" which marks the beginning of a
new era of highly contextualized and
personalized experiences, delivered when and
where we want them. AIAIAIAIAI belongs to one of that
theme. The thematic content as mentioned in
mobile world congress (MWC) site is stated as:
"With a market projected to reach $70 billion by
2020, artificial intelligence is poised to have a
transformative effect on consumers, enterprises,
and governments around the world. AI explores the
real potential of artificial intelligence, as well as how
we manage such a profound technological revolution
and its impact on our professional and personal lives".
[1]
The telecom operators are facing new
opportunities and challenges with the exponential
growth of IoT devices and 4G/5G network data, the
acceleration of cloud-based network adoption, the
convergence of OTT, and the increasing expectations
on quality of service and customer experience. Many
business opportunities are coming from the gold mine
of data generated by customers who are connecting
to their networks through a variety of applications at
a faster rate than ever. Telecom service providers can
now collect data at every step of a subscriber’s
journey: from detailed device, network, and operations
data and web and mobile applications to geo-location,
customer care and consumer profile data, call detail
record (CDR), service usage, and billing data.
The challenges reside in handling higher
volumes of data and extracting actionable insights
while improving network efficiencies and lowering
operational expenses. This is where, AI can help
the telecommunication industry, by opening the
door for: [2]
Data monetization: Utilizing subscribers’ data
to sell additional services and create new ones thus
bringing in new revenue streams
Intelligent network deployments: Using
automation to efficiently operate networks and save
costs
Fig. Applications of AI in telecom sectors
Some of the potential application fields of AI in
Telco are described as below:
1. Customer experience & relationship1. Customer experience & relationship1. Customer experience & relationship1. Customer experience & relationship1. Customer experience & relationship
management:management:management:management:management:
AI tools can help in better customer experience
and better customer relationship management (CRM).
For example, using AI powered chatbot instead of
IVR based customer contact center will have far more
advantages and features. Chatbots are AI-based
conversation agents that are being used in many
different customer-engagement scenarios. Such
chatbot will be intelligent enough to answer/solve
customer queries itself immediately, it will be
available 24 * 7, less human intervention will be
required. This eliminates frustrating delays and errors
in customer service, particularly for handling
customer complaints. For example,
ada.support chatbot (https://ada.support/), avamo
bots of telecommunications
Avaamo telecommunication chatbot (https://
www.avaamo.com/telecommunications/) are
some of the examples of chatbots that can be used
in customer support center.
AI-enabled CRMs help companies assess which
136
15th
Anniversary Souvenir 2019
kGw|f“} jflif{sf]T;j :dfl/sf @)&%
customers could be the
most profitable and likely
to respond to sales
outreach. It helps in
bringing different service
packages based on
particular group of users,
geolocation. Adding an
AI model to the
telecom’s CRM can
make the salesperson see
potential offer
recommendations as the
customer is pulled up in
the CRM software.
2. Network security2. Network security2. Network security2. Network security2. Network security
With the rapid
increase in the large
number of connected
devices, their increase in intelligence level, their
compute power, lines of codes and wireless
connections with the outside world is making them
attractive victim for cyber attacker. AI has the
capability to process and analyze millions of data
points at high speed, scan systems and networks for
vulnerabilities, learn to detect and mitigate anomalies
or suspicious behavior, identify patterns of new
security threats and block suspicious traffic in real
time. Intel’s Apache Spot [3] , a community-driven
cyber security project based on AI, is a great example.
3. Predictive maintenance3. Predictive maintenance3. Predictive maintenance3. Predictive maintenance3. Predictive maintenance
For telecom service providers, predictive
maintenance is essential for providing continuous and
reliable solutions to the customers. For example, it is
needed to keep a continuous watch on their
infrastructure and equipment, from cell towers, power
lines, network routers/switches/firewalls and heavy
equipment to servers in data centers so that they can
ensure a reliable and secure network. Any downtime
anywhere along the line can be extremely costly and
is related with goodwill of the company. AI systems
can monitor the state of equipment, identify patterns
that predict failure and perform maintenance on a
preemptive basis. In addition, quality control through
image analysis can detect if a field technician has
missed something. AT&T takes image analysis further
by planning to use drones [4] to capture video of cell
towers, which the algorithms process to understand
if there are any issues with the infrastructure.
Mazin Gilbert, VP of Advanced Technology at
AT&T Labs predicts that predictive network
maintenance will continue to drive favorable expense
trends over the next several years.
"We are implementing AI to help us to identify
where these breakpoints are, and help to repair those
in an automated way without human intervention.
This goes for hardware failure, software failures."
– Mazin Gilbert, VP of Advanced Technology
at AT&T Labs [5]
4. Revenue Assurance4. Revenue Assurance4. Revenue Assurance4. Revenue Assurance4. Revenue Assurance
Revenue data comes from dozens of databases,
about millions of customers, and through potentially
hundreds of partners. While logic is applied to the
databases, if something breaks, or there’s an incident
of fraud, the company could lose thousands to
millions of amount. Anomaly detection algorithms
find deviances from usual in the trends. The model
then alerts the experts, who evaluate whether the
deviation is expected, for reasons like changes in the
pricing policy or pricing specials, or there’s an issue
with the system, or there’s potential for fraud. [6]
As the algorithm runs and the experts provide
feedback, it learns and gets better at understanding
the anomalies and detecting fraud.
So, time has come for Nepal Telcom also to think
and bring Artificial Intelligence (AI) based products
and solutions in this competitive market, for better
quality of service, and customer satisfaction.
ReferencesReferencesReferencesReferencesReferences
[1] "Mobile world congress 2019," [Online]. Available:
https://www.mwcbarcelona.com/about/event-themes/.
[2] "Intel AI," [Online]. Available: https://ai.intel.com/
taking-telecom-new-heights- artificial-intelligence/
#gs.ddlSL2g.
[3] "Apache spot," [Online]. Available: https://ai.intel.com/
taking-telecom-new-heights-artificial-intelligence/
#gs.ddlSL2g.
[4] "Industrial Uses of Drones – 5 Current Business
Applications," December 2018. [Online]. Available:
https://emerj.com/ai-sector-overviews/industrial-uses-of-
drones-applications/.
[5] "Artificial Intelligence in Telecom Business," [Online].
Available: [https://www.mindtitan.com/case/ai-in-
telecom-business/.
[6] "Taking Telecom to New Heights with Artificial
Intelligence," February 2018. [Online]. Available: [https:/
/ai.intel.com/taking-telecom-new-heights-artificial-
intelligence/].
Fig. Typical chatbot for
customer contact center

Introducing AI into telcos

  • 1.
    134 15th Anniversary Souvenir 2019 kGw|f“}jflif{sf]T;j :dfl/sf @)&% Shyam Krishna Khadka Engineer ISSD, Nepal Telecom Introducing Artificial Intelligence into Telecom Companies (Telco) The telecom operators are facing newThe telecom operators are facing newThe telecom operators are facing newThe telecom operators are facing newThe telecom operators are facing new opportunities and challenges with theopportunities and challenges with theopportunities and challenges with theopportunities and challenges with theopportunities and challenges with the exponential growth of IoT devices and 4G/exponential growth of IoT devices and 4G/exponential growth of IoT devices and 4G/exponential growth of IoT devices and 4G/exponential growth of IoT devices and 4G/ 5G network data. The challenges reside in5G network data. The challenges reside in5G network data. The challenges reside in5G network data. The challenges reside in5G network data. The challenges reside in handling higher volumes of data and extractinghandling higher volumes of data and extractinghandling higher volumes of data and extractinghandling higher volumes of data and extractinghandling higher volumes of data and extracting actionable insights while improving networkactionable insights while improving networkactionable insights while improving networkactionable insights while improving networkactionable insights while improving network efficiencies and lowering operationalefficiencies and lowering operationalefficiencies and lowering operationalefficiencies and lowering operationalefficiencies and lowering operational expenses. This is where, AI can help theexpenses. This is where, AI can help theexpenses. This is where, AI can help theexpenses. This is where, AI can help theexpenses. This is where, AI can help the telecommunication industrytelecommunication industrytelecommunication industrytelecommunication industrytelecommunication industry What is AI?What is AI?What is AI?What is AI?What is AI? Artificial Intelligence (AI): Artificial + Intelligence. By word, AI means bringing human like intelligence to machines, making machines think and behave like human. The works which are tedious, repetitive, requiring large human population can now be simplified by using AI techniques. Artificial Intelligence (AI) has become one of the hottest technologies that are being used in various fields like agriculture, industrial process, security, network, telecommunication, health, biology etc. Some of the examples of AI in our real life include: 1. Email Filters in Gmail1. Email Filters in Gmail1. Email Filters in Gmail1. Email Filters in Gmail1. Email Filters in Gmail 2. Chatbots2. Chatbots2. Chatbots2. Chatbots2. Chatbots Chatbots recognize words and phrases in order to (hopefully) deliver helpful content to customers who have common questions. Sometimes, chatbots are so accurate that it seems as if you’re talking to a real person. 3. Navigation and travel apps3. Navigation and travel apps3. Navigation and travel apps3. Navigation and travel apps3. Navigation and travel apps AI is used in apps like Google/Apple maps for navigating, or calling an Uber, or booking a flight ticket. Both Google and Apple along with other navigation services use artificial intelligence to interpret hundreds of thousands of data point that they receive to give you real-time traffic data. When we are calling an Uber, both the pricing and the car that matches our ride request are decided by AI. As we can see, AI plays a significant role in how we reach from point A to point B. 4. Product Recommendation systems4. Product Recommendation systems4. Product Recommendation systems4. Product Recommendation systems4. Product Recommendation systems Amazon and other online retailers use AI to gather information about your preferences and buying habits. Then, they personalize our shopping experience by suggesting new products tailored to our habits. 5.5.5.5.5. Fraud detection in financial sectorsFraud detection in financial sectorsFraud detection in financial sectorsFraud detection in financial sectorsFraud detection in financial sectors There are many situations in financial sectors that AI powered software/technologies are used in fraud detection. For example: Getting SMS/Email alert if an unusually large transaction takes place in your account and many others. The underlying working principle of AI lies on data. AI needs huge amount of data, to work. AI techniques have capacity to analyze huge data, detect patterns among them, classify and predict them. For a Telco like Nepal Telecom, we have large data that may be either of customers or of network elements (NE) data or employee data. Depending on the type of data, AI can be used in corresponding fields. Potential application fields of AI in telecomsPotential application fields of AI in telecomsPotential application fields of AI in telecomsPotential application fields of AI in telecomsPotential application fields of AI in telecoms In mobile congress conference that was held at Barcelona, Spain in 2018, "Applied AIApplied AIApplied AIApplied AIApplied AI" was one of
  • 2.
    135 15th Anniversary Souvenir 2019 kGw|f“}jflif{sf]T;j :dfl/sf @)&% the eight themes of the conference. The intention was to help the industry cut through the complexity of artificial intelligence (AI). The same trend regarding AI has been continued in this year of mobile congress committee too (going to be held in February 2019). The event theme is "IntelligentIntelligentIntelligentIntelligentIntelligent ConnectivityConnectivityConnectivityConnectivityConnectivity" which marks the beginning of a new era of highly contextualized and personalized experiences, delivered when and where we want them. AIAIAIAIAI belongs to one of that theme. The thematic content as mentioned in mobile world congress (MWC) site is stated as: "With a market projected to reach $70 billion by 2020, artificial intelligence is poised to have a transformative effect on consumers, enterprises, and governments around the world. AI explores the real potential of artificial intelligence, as well as how we manage such a profound technological revolution and its impact on our professional and personal lives". [1] The telecom operators are facing new opportunities and challenges with the exponential growth of IoT devices and 4G/5G network data, the acceleration of cloud-based network adoption, the convergence of OTT, and the increasing expectations on quality of service and customer experience. Many business opportunities are coming from the gold mine of data generated by customers who are connecting to their networks through a variety of applications at a faster rate than ever. Telecom service providers can now collect data at every step of a subscriber’s journey: from detailed device, network, and operations data and web and mobile applications to geo-location, customer care and consumer profile data, call detail record (CDR), service usage, and billing data. The challenges reside in handling higher volumes of data and extracting actionable insights while improving network efficiencies and lowering operational expenses. This is where, AI can help the telecommunication industry, by opening the door for: [2] Data monetization: Utilizing subscribers’ data to sell additional services and create new ones thus bringing in new revenue streams Intelligent network deployments: Using automation to efficiently operate networks and save costs Fig. Applications of AI in telecom sectors Some of the potential application fields of AI in Telco are described as below: 1. Customer experience & relationship1. Customer experience & relationship1. Customer experience & relationship1. Customer experience & relationship1. Customer experience & relationship management:management:management:management:management: AI tools can help in better customer experience and better customer relationship management (CRM). For example, using AI powered chatbot instead of IVR based customer contact center will have far more advantages and features. Chatbots are AI-based conversation agents that are being used in many different customer-engagement scenarios. Such chatbot will be intelligent enough to answer/solve customer queries itself immediately, it will be available 24 * 7, less human intervention will be required. This eliminates frustrating delays and errors in customer service, particularly for handling customer complaints. For example, ada.support chatbot (https://ada.support/), avamo bots of telecommunications Avaamo telecommunication chatbot (https:// www.avaamo.com/telecommunications/) are some of the examples of chatbots that can be used in customer support center. AI-enabled CRMs help companies assess which
  • 3.
    136 15th Anniversary Souvenir 2019 kGw|f“}jflif{sf]T;j :dfl/sf @)&% customers could be the most profitable and likely to respond to sales outreach. It helps in bringing different service packages based on particular group of users, geolocation. Adding an AI model to the telecom’s CRM can make the salesperson see potential offer recommendations as the customer is pulled up in the CRM software. 2. Network security2. Network security2. Network security2. Network security2. Network security With the rapid increase in the large number of connected devices, their increase in intelligence level, their compute power, lines of codes and wireless connections with the outside world is making them attractive victim for cyber attacker. AI has the capability to process and analyze millions of data points at high speed, scan systems and networks for vulnerabilities, learn to detect and mitigate anomalies or suspicious behavior, identify patterns of new security threats and block suspicious traffic in real time. Intel’s Apache Spot [3] , a community-driven cyber security project based on AI, is a great example. 3. Predictive maintenance3. Predictive maintenance3. Predictive maintenance3. Predictive maintenance3. Predictive maintenance For telecom service providers, predictive maintenance is essential for providing continuous and reliable solutions to the customers. For example, it is needed to keep a continuous watch on their infrastructure and equipment, from cell towers, power lines, network routers/switches/firewalls and heavy equipment to servers in data centers so that they can ensure a reliable and secure network. Any downtime anywhere along the line can be extremely costly and is related with goodwill of the company. AI systems can monitor the state of equipment, identify patterns that predict failure and perform maintenance on a preemptive basis. In addition, quality control through image analysis can detect if a field technician has missed something. AT&T takes image analysis further by planning to use drones [4] to capture video of cell towers, which the algorithms process to understand if there are any issues with the infrastructure. Mazin Gilbert, VP of Advanced Technology at AT&T Labs predicts that predictive network maintenance will continue to drive favorable expense trends over the next several years. "We are implementing AI to help us to identify where these breakpoints are, and help to repair those in an automated way without human intervention. This goes for hardware failure, software failures." – Mazin Gilbert, VP of Advanced Technology at AT&T Labs [5] 4. Revenue Assurance4. Revenue Assurance4. Revenue Assurance4. Revenue Assurance4. Revenue Assurance Revenue data comes from dozens of databases, about millions of customers, and through potentially hundreds of partners. While logic is applied to the databases, if something breaks, or there’s an incident of fraud, the company could lose thousands to millions of amount. Anomaly detection algorithms find deviances from usual in the trends. The model then alerts the experts, who evaluate whether the deviation is expected, for reasons like changes in the pricing policy or pricing specials, or there’s an issue with the system, or there’s potential for fraud. [6] As the algorithm runs and the experts provide feedback, it learns and gets better at understanding the anomalies and detecting fraud. So, time has come for Nepal Telcom also to think and bring Artificial Intelligence (AI) based products and solutions in this competitive market, for better quality of service, and customer satisfaction. ReferencesReferencesReferencesReferencesReferences [1] "Mobile world congress 2019," [Online]. Available: https://www.mwcbarcelona.com/about/event-themes/. [2] "Intel AI," [Online]. Available: https://ai.intel.com/ taking-telecom-new-heights- artificial-intelligence/ #gs.ddlSL2g. [3] "Apache spot," [Online]. Available: https://ai.intel.com/ taking-telecom-new-heights-artificial-intelligence/ #gs.ddlSL2g. [4] "Industrial Uses of Drones – 5 Current Business Applications," December 2018. [Online]. Available: https://emerj.com/ai-sector-overviews/industrial-uses-of- drones-applications/. [5] "Artificial Intelligence in Telecom Business," [Online]. Available: [https://www.mindtitan.com/case/ai-in- telecom-business/. [6] "Taking Telecom to New Heights with Artificial Intelligence," February 2018. [Online]. Available: [https:/ /ai.intel.com/taking-telecom-new-heights-artificial- intelligence/]. Fig. Typical chatbot for customer contact center