Preparing for AI and generative AI: essential strategies for telcos

AI is becoming a commercial imperative for the highly competitive, fast evolving telecommunications industry. In the first of a two-part blog series, we made the case for swift adoption within the sector, because of the technology’s transformative impact on operational efficiency, revenue generation and customer experience.

However, according to the global industry association TM Forum, 45% of telcos have yet to exploit AI effectively, let alone generative AI (gen AI), and many don’t have a clear adoption strategy. This must change; average revenues for the top 10 telecoms operators worldwide grew by just 2.7% per year from 2017-2023.

Telcos need to focus on building the right foundation for a scalable gen AI operating model with maximum impact across the entire business. This should be innovative and holistic, based on a centralised strategy, healthy data, a talent ecosystem and the right mindset.

Putting the building blocks in place: a centralised approach

The most successful gen AI strategy is centrally led, with a strong vision aligned to business objectives and a clear focus on creating value for each business unit within the organisation. This is the fastest, most efficient approach to building gen AI capability at scale, as a centralised team orchestrates technological innovation and demonstrates proof of value (or not) for a variety of use cases. Over half of telcos have already set up a centre of excellence for gen AI, according to TM Forum, and they’re beginning to appoint CXOs for AI.

This centralised approach must also be inclusive, building trust with the business to secure buy-in and align with their key objectives. CIOs need to understand each business unit’s goals, share all data and results and show the positive impact of gen AI as key drivers for strengthening trust and increase the rate of experimentation. Robust governance programme is also vital, to ensure that gen AI is used responsibly, safely and securely.

A healthy data pipeline

There’s arguably no industry better placed to harness the power of gen AI than telecommunications. Data is this technology’s lifeblood and telcos hold vast quantities of it, from their extensive networks, their operations and from their customers. Data sets must be high-quality and relevant to ensure accurate LLM outputs, however, and here lies a challenge.

Almost half of top-tier telcos rank data collection as the greatest impediment to developing AI use cases, according to a survey by Nokia and Analysys Mason. Barriers include siloed, fragmented and unclean data, stored in different formats and often unlabelled.

Before embarking on the gen AI journey, telcos urgently need to modernise and consolidate their data architectures. Robust governance is vital to ensure that all data is accurate, secure and meets privacy requirements.

Creating a talent ecosystem

Most telcos are likely to buy in their gen AI models rather than build from scratch, because of the cost and complexity of designing a bespoke system. Adopting off-the-shelf Large Language Models (LLMs) is quicker and more cost efficient, but their highly generalised applications don’t meet the sector’s specific needs and there’s a dearth of in-house expertise to train them. What’s the answer?

Strategic partnerships bring the external expertise and resources which telcos need to plug the skills gap. Together, in-house teams and partners form a gen AI ecosystem which empowers telcos to leverage LLMs for customised solutions. Using telcos’ rich data, partners can fine-tune foundation models so that gen AI is successfully embedded in telecom-specific workflows. This collaborative arrangement ensures that the right talent is in place to unlock the full potential of gen AI.

High-impact deployment

As telcos look to scale gen AI capability throughout their organisations, they should consider how the technology fits within their existing architecture and where it provides greatest value. An obvious area of focus is software engineering, which powers every aspect of the telecom sector, from network operations to customer support.

Cognizant has two solutions for adopting and accelerating gen AI transformation within software development and operations:

• Flowsource is a full-stack platform which integrates gen AI and automates workflows at every stage of the software development lifecycle, delivering high-quality code quickly and securely. 
• Neuro AI, a platform which bundles tools, models and frameworks to empower enterprise-grade gen AI adoption at speed, flexibly, securely and responsibly.

Together, these two platforms empower high-impact deployment of gen AI in the core area of software engineering, providing consistency, transparency and all the necessary guardrails. From this basis, telcos can build momentum by rolling out gen AI to other areas of the business.

Mindset

Technology is changing fast and telcos face multiple challenges in embracing gen AI. They must remain curious and open-minded to enable this transformative technology to reach and revolutionise every corner of the organisation. However, there will inevitably be teething problems.

Telcos are absorbed in the “five nines”, or the availability of network services for 99.999% of the time. Whilst this is a fantastic service level target for their customers, they’re not going to achieve “five nines” in gen AI deployment. So it’s time for a mindset shift; they must experiment, learn from any mistakes and build that learning into the next gen AI iteration.

With this attitude, telcos can unlock the exciting potential of gen AI. The roadmap for success includes a centralised strategy, high-quality data and robust governance to ensure that the technology is used safely and responsibly. Telcos can also make strategic partnerships to attract talent and adapt gen AI to the sector-specific demands.

Starting with “quick wins” – the use cases which can be introduced rapidly and effectively to start the gen AI transformation – telcos can build on these successes to continue their gen AI journey with confidence and gain the competitive advantage.

If you would like to learn more, please get in touch.