Data Analytics & AI

Make it happen – Strategies for adopting AI technology

This is an extract from our new report on AI, AI: The time is now, download the report for the full insight.

Most large communications service providers (CSPs) are taking the first tentative steps toward embracing artificial intelligence (AI) and machine learning, but being proactive, rather than waiting for technology providers to offer ideas and strategies, can be difficult.

Today CSPs procure the vast majority of their IT and network technology from suppliers that focus specifically on the telecom industry and conform to agreed standards, but those companies aren’t necessarily the best sources of AI technology. The idea of recruiting data scientists and using open source technologies to develop internal AI expertise is also intriguing, but many CSPs lack the competence or confidence to invest organically in AI.

Thousands of telecoms executives are gathering at Digital Transformation World in Nice, to explore, collaborate, and become. Discover which of the ten AI Catalyst projects getting ready for Digital Transformation World is the best and contribute to the discussion about how AI fits into the Open Digital Architecture of the future. 

Despite the challenges, the time is now for operators large and small to begin adopting AI and machine learning. The technologies are necessary to remain competitive, support the growing internet of things (IoT) and deliver the kinds of experiences customers are demanding. Here are some steps CSPs can take to get started with AI or to progress nascent programs:

Empower a leader

It’s not just about understanding AI; it’s also about working across the organization to create use cases and identify points of commonality. A leader in the organization must be empowered to do this – likely either a CIO or chief data officer.


Whether CSPs want to develop their own AI platform and tools using open source software or are more inclined to partner with suppliers, they must educate themselves about the technology and the skills and resources required to leverage AI’s potential. This can include experimenting with AI systems, seeking out partners, and joining relevant standards bodies and open source groups.

Integrate AI with analytics

AI and machine learning need data. Indeed, AI enables CSPs to act upon the data they collect and store, so operators should view AI as an evolution of their data strategy.

To facilitate the deployment of AI use cases it is imperative that CSPs natively expose data, remove it from silos and make it accessible via APIs throughout the entire organization. They should plan to create centralized data lakes to do this.


To attract the interest of leading AI technology suppliers, CSPs must be easy to do business with. They need to standardize the data collection and management – without prejudicing or prejudging the use cases that they might want to develop. This will allow technology partners to find a common approach to plugging CSPs’ data into their platforms.

Be realistic

The market for AI talent is red hot and CSPs are not the obvious career choice for a computer scientist who wants to build a career in AI. Even if they are able to recruit a team of specialists, it’s unlikely they will be able to keep pace with demand for AI expertise across the organization.

The siloed structures and rigid processes in many CSPs mean that ambitious AI researchers are likely to get frustrated in their roles. Unless chief strategy officers can address pay-scales and create interesting roles for AI specialists, operators may need to accept that they will have to partner with vendors and systems integrators.

Think automation

The early excitement around network functions virtualization (NFV) and software-defined networking (SDN) has waned even though CSPs recognize that ultimately their networks will be virtualized and cloud-based. They recognize that unless networks can be automated, the deployment of NFV and SDN may actually increase operational costs. AI and machine learning are key to advanced automation.

Keep your options open

Many of the CIOs and chief architects we spoke with are trying to decide what sort of structure or centralized capability to build around AI. In these exploratory stages it is important for CSPs to incubate AI projects across the business without necessarily pre-judging how and when they should come together, and the extent to which a central technology capability, platform or repository should be adopted.

However, we do believe that centralized coordination is needed to address data management, governance, warehousing and data lakes, and common analytical and AI technologies.

Be a guinea pig

Many technology suppliers want to showcase their AI capabilities. This represents a good opportunity for CSPs to gain useful knowledge and potentially derive business benefits from projects that are sponsored by third parties.


AI is an area where there is everything to gain and very little to lose from collaborating with other CSPs. Industry bodies such as TM Forum, ETSI and the Facebook-backed Telecoms Infra Project, are all establishing AI initiatives, and we urge CSPs to participate as widely as possible.


    About The Author

    Chief Analyst

    Mark Newman is an analyst with 25 years of experience delivering insights on the future of the telecoms sector to senior level executives and audiences. Mark’s recent research has focussed on telecoms operator business models, digital transformation, service provider diversification, and the intersection between Internet and telecoms. He delivers analysis, presentations, strategy sessions and workshops to global audiences, helping them to plan for the changes that technology and disruptive new business models that will fundamentally transform their businesses. Mark was Chief Research Officer at Informa Telecoms & Media and Ovum before leaving to set up his own research firm, ConnectivityX, in 2016. He joined the TM Forum as Chief Analyst in February 2017.

    Leave A Reply

    Back to top