From the Foothills to the Summit
Scaling agentic AI for the autonomous Telco
The agentic AI Telco of the future is right around the corner. It’s not a question of if it happens, but how soon. Those who act now will be in pole position to capitalize on the opportunities it offers; those that delay risk falling behind those who have put AI agents at the heart of their business. The predicted unlocking of $60bn in cost savings is just the tip of the iceberg.
Telcos currently exploring the potential of agentic AI must realize that it represents a paradigm shift, not only in the use of data and technology, but in the core operations of the organization. Agentic AI represents much more than another technological wave, as it is ushering in a wholesale transformation of every aspect of the business. Deploying AI agents at enterprise scale will demand fundamental changes to business processes, operations, and platforms but will also transform the growth and commercial direction of telcos.
Yesterday’s tomorrow
This future is already close at hand. Over the past few blogs, I have shared how AI agents are being deployed in customer experience, fraud and cyber security, business operations and network operations. There are powerful examples of improving outcomes, increasing productivity, and saving money, and they highlight a sensible “modular” approach to building AI agent activity. Selecting high-impact areas of the business where agents can be relatively quickly deployed to deliver rapid return on investment is an effective way to prove the concept for wider implementation.
But these are just the foothills of the real transformation. Recent reports suggest that these early projects are the basecamps from which organizations begin the climb toward fully autonomous enterprises, where human and machine labor work hand in hand to be more productive, more innovative, faster, and safer than before. Network management, customer experience and productivity represent the most likely routes to the summit.
A recent Omdia report shows that 48% of respondents expect the initial impact of agentic AI to be in improving customer engagement. By moving beyond recommending solutions to making decisions and automatically implementing changes in real time, agentic AI will radically enhance customer experience. But the same report also found that 41% of communications service providers (CSPs) saw network management as the most significant area of impact.
Interestingly, several experts see significant impact not only in improving productivity of human workers through “copilots” that can help with coding or creating content, but also through removing friction from internal processes. Boston Consulting Group finds that AI-powered workflows can accelerate business processes by 30% to 50% in areas ranging from finance and procurement to customer operations.
Leading from the front
The journey to enterprise-wide agentic AI must be led by CEOs with the vision to create new business models and processes that embrace its power to do things differently, and more fundamentally, to do different things. In the memorable words of Brent Collins at Intel, this is not about “paving the cow path.” Instead, leaders need to be bold in recreating their organisations around AI agents. McKinsey recently found that high-performing organizations were almost three-times more likely to report that they plan to “fundamentally redesign their workflows” with AI. McKinsey also notes that these companies are much more likely to have scaled AI deployments beyond experimentation and proof of concept stage.
Achieving this transformation at scale depends on having the right data, the right people, and the right platforms in place. According to AI leaders surveyed by Deloitte, the top primary challenge for adopting agentic AI is integrating with legacy systems. Gartner predicts that over 40% of agentic AI projects will fail by 2027 because legacy infrastructure lacks the real-time execution capability and modern APIs needed for agent integration. Enterprises therefore need hybrid architectures that combine deterministic, rule-based systems with agentic AI. These architectures enable agents to search across enterprise data and provide the foundation for AI memory that supports contextual intelligence.
Leaders must also envision a deep-rooted shift in their workforce, from purely human, to human and silicon workers collaborating. Successfully scaling agentic AI may mean treating agents as a new form of digital labor that complements human workers. Human roles may transform not only away from the mundane and repetitive, but from execution towards planning, strategy, governance, and innovation.
Get to know your digital teammates
Critically, the success of agentic AI does not lie only in the hands of a few data scientists, but includes all roles in multi-functional, enterprise-wide teams. Everyone will need to know not only how to use AI agents, but how to trust them, and how to work with digital teammates. Only then can the opportunity to transform, or even replace, existing business, operational and technical processes be managed in a way that creates win-win situations for the telco, its employees, and customers.
As this “silicon workforce” expands more and more tasks will be automated. Some CSPs are already launching registries of pre-trained agents, like recruitment centers that allow agents to be employed “out of the box” to deliver specific roles. As these proliferate, the number, variety, and frequency of agent-to-agent communications will explode, with human roles potentially managing agents that manage their own teams of agents to deliver prescribed outcomes. We can already see the potential for agents to define and deploy new agents autonomously within human-established governance frameworks. Emerging open standards such as Model Context Protocol (MCP) and Agent-to-Agent Protocol (A2A) will be vital to manage these operations.
The new digital transformation
To successfully navigate these complexities, CSPs must establish a trusted, scalable, and explainable data infrastructure. Agentic AI systems rely on anchors like deterministic machine learning models and ontologies to root their reasoning in reliable truths, enhancing contextual awareness and avoiding erratic behaviour. Teradata’s AI + Knowledge Platform provides the foundation necessary for true enterprise autonomy. Our architecture enables organizations to ground every AI innovation in enterprise knowledge, ensuring performance at scale. By integrating capabilities like RAG and supporting emerging standards such as MCP and A2A, the Teradata platform empowers enterprises to unify, analyze, and act on structured and unstructured data, delivering trusted autonomy that is prepared for every new challenge.
Agentic AI will be a strategic compass for the telecom industry. Leaders who move early and invest deliberately can translate these capabilities into measurable gains in efficiency, network performance, and customer experience. Organizations that delay may find it harder to compete as faster, more automated operating models become the norm. Get in touch now to learn how Teradata can help scale Agentic AI effectively in your organization.