Market Barriers & Market Overview

Telecom operators are transforming their businesses, but that transformation is slow.

Most telecom operators are hybrid/patchwork organizations. Most of the larger existing mobile operators

are the result of numerous acquisitions. Consequently, they use various legacy BSS/OSS) solutions and

networks, because ripping them out and replacing them with sleeker, more uniform systems would be too

costly. The majority of the money that telecom operators will invest in AI will be triggered when newer

systems based on SDNs and/or NFV are in place.

SDNs and NFV have been talked about since 2012 by telecom operators as their new reality for agile

programmable networks, but the rollout of these technologies for telecom operators has been slow.

Subsequently, AI-driven solutions for network monitoring and operations and the drive for autonomous

networks designed for programmable networks have been hindered.

Affordable and reliable power supply is a challenge for many CSPs in Africa. Telecom AI use cases need

significant power requirements, which could limit the use of AI to fully-functional data centers and cloudbased solutions.

Globally, an immature ecosystem for telecom AI use cases has formed, made up of legacy telecom network

and BSS/OSS vendors; broad-based automated customer service specialists; CRM providers; open-source

communities and organizations; established cybersecurity vendors; and a small but impressive number of

startups. Some AI solutions, particularly for fraud management and cybersecurity, have been long

established with telecom operators. Customer service and marketing VDAs took root with telecom operators

in late 2016 and their use is spreading rapidly. For the most complex issue facing telecom operators—

network operations management—it is very early days for AI solutions, though legacy vendors and some

bold startups, as well as operator-led self-built solutions, are beginning to move forward.