Image credit: qtxasset.com

The Shifting Landscape of Telecom AI Development

In recent years, telecommunications companies (telcos) are taking an increasingly hands-on approach to artificial intelligence (AI) development. Those who have traditionally relied on off-the-shelf solutions are now investing in building their own models from the ground up. This so-called DIY (do-it-yourself) trend is proving to be both bold and risky. By developing internal teams of data scientists, AI experts, and engineers, telcos aim to tailor AI solutions that meet their unique business needs.

The fundamental question arises: Why are telcos moving towards a DIY approach when proven commercial solutions are available? A unique combination of customization needs, differentiation from competitors, and a desire for control has driven this trend. Despite many challenges, the benefits of a DIY model are attractive to companies eager to gain a competitive edge and build specialized capabilities.

Benefits and Challenges of the DIY Approach

While the ROI potential of creating bespoke AI solutions can be high, it does not come without significant challenges. Developing AI systems in-house demands extensive resources, not only financial but also time and expertise. Telcos need to build multi-disciplinary teams that can integrate AI models with their existing infrastructure. This includes a deep understanding of data engineering, machine learning, and operational demands.

For many organizations, this can lead to delays and distractions. The internal focus on building AI tools takes valuable talent away from other critical initiatives. According to McKinsey, continued reliance on a DIY model can inadvertently slow down innovation within the industry. The pressure to maintain competitiveness drives telcos to adopt AI strategies that may not align with their overarching business goals.

Mapping the Road Ahead: When to Build vs. Buy

Moving forward, telcos need to strategically assess when to build their own AI models and when to leverage existing commercial solutions. This balancing act requires a clear-eyed evaluation of resources and overall business objectives. Collaboration with gen AI solution providers can offer the scaffolding necessary to enhance internal capabilities without sacrificing innovation.

For instance, Deutsche Telekom's Ask Magenta chatbot serves as an illustrative case of how the blending of DIY and commercial solutions can yield significant results. By utilizing a large language model (LLM) within its operations, Deutsche Telekom has enhanced user experience through more accurate responses, showcasing that adopting existing technology can complement bespoke development.

Developing a Holistic AI Strategy

The successful implementation of AI within telcos requires a well-thought-out approach that encompasses not just the technological side but a comprehensive vision aligned to business goals. Building a robust business-led roadmap for AI initiatives is crucial. This means evaluating the roadblocks, laying out priorities, and determining the best approach for each project — whether that involves a DIY solution or a partnership with vendors.

Telcos also need to invest in talent acquisition and training. Specialists in data science and machine learning should be incorporated into strategic planning discussions. A skilled workforce combined with sound strategy can lead to effective AI implementation, bridging the gap between emerging technologies and real-world application.

In addition, scaling operations as AI technologies mature plays a significant role. Telcos must ensure that they have the right support system, from technical resources to organizational buy-in, to move past pilot stages into fully functioning solutions that deliver measurable business value.