Telecom operators are moving from generative AI experimentation to agentic AI-autonomous systems that reason, make context-aware decisions, and act within defined governance frameworks. This shift, detailed by Amdocs Division President Samit Banerjee, lets service providers orchestrate intelligent agents across customer care, network operations, and enterprise workflows without ripping out existing BSS/OSS infrastructure. For operations leaders, the change means faster resolution times, predictive network management, and a fundamental redesign of how work gets done.
What separates agentic AI from traditional automation
Traditional automation follows predefined rules. Generative AI creates content and assists decisions. Agentic AI does both and more-it monitors events, collaborates with other agents, and executes complex workflows with minimal manual intervention. "Instead of isolated AI use cases, operators can orchestrate intelligent agents across customer care, network operations, service assurance, sales, and enterprise functions," Banerjee said. The goal is not replacing people. It is augmenting human expertise with AI that acts at scale, keeping human oversight on decisions that matter.
Telecom networks generate enormous data volumes in highly dynamic environments. Agentic AI enables a shift from reactive to predictive and autonomous operations. Operators see faster resolution, better service quality, and employees freed for higher-value work. Banerjee emphasized that governance and human judgment remain central: "The operators that win will be fast, adaptable, and disciplined about turning intelligence into measurable results; with governance and human judgment still in the loop."
Data foundations determine whether AI scales or stalls
AI is only as good as the data behind it. Operators hold some of the richest data in any industry-network performance, customer interactions, billing, service assurance, and operational intelligence-but it is often fragmented across systems. GSMA's 2026 report, Scaling Telco AI, identifies data readiness as a key blocker. Three priorities stand out: high-quality, governed data with clear ownership; modernizing the stack so AI can reach enterprise-wide data without creating new silos; and embedding AI into end-to-end workflows tied to measurable KPIs like resolution time, cost, CSAT, and revenue.
Banerjee put it plainly: "No trust in the data, no consistent outcomes." Treating data as a business asset rather than an IT line item puts operators ahead of most of the market. For operations professionals building AI for Operations capabilities, this means data governance is not a prerequisite to tick off-it is the foundation that determines whether AI projects deliver returns or stall after the pilot phase.
How AI rewires customer experience and revenue models
Efficiency gains matter, but AI's larger impact comes from delivering personalized, proactive experiences. Tailored offers, predicting issues before customers encounter them, real-time network optimization, and contextual support across channels are now possible. What is shifting fundamentally is the traditional sales funnel. The old model of moving customers stage by stage through awareness, consideration, and purchase is breaking down. AI allows all three to happen in one continuous conversation.
"The job isn't designing content per funnel stage anymore; it's designing the conversation itself: how it sounds, how it adapts, whether it feels like the same brand from start to finish," Banerjee said. He calls this personality engineering-giving an AI agent a consistent character so the brand feels the same whether a customer is browsing, comparing, or buying. Customers never see the agent, but they feel whether it understands them. That increasingly decides loyalty.
On the business side, AI opens monetization opportunities for enterprise customers, faster service launches, and differentiated digital experiences. Operators can move from connectivity providers to digital service partners supporting smart cities, healthcare, and other sectors. Competitive advantage will not come from connectivity alone, but from how intelligently AI is used to create new business models.
Orchestrating agents across existing infrastructure
The real barrier to enterprise AI is not model availability-it is operationalizing AI across complex telecom environments without disrupting existing investments. Amdocs' aOS platform functions as an orchestration layer that lets specialized agents collaborate across customer care, network operations, service assurance, monetization, and enterprise workflows. Built on an open, telco-specific cognitive core, aOS combines domain expertise, pre-built agents, multi-LLM support, and open APIs.
It integrates with existing BSS/OSS systems, so modernization happens incrementally. Operators embed intelligence into existing processes, orchestrate agents across domains, and automate high-impact workflows with governance and human oversight intact. For operations managers following an AI Learning Path for Operations Managers, the architecture lesson is clear: the winning approach layers intelligence on top of current systems rather than demanding a disruptive rip-and-replace.
Why this matters for operations leaders
Agentic AI changes the operations playbook in three concrete ways. First, it shifts teams from reactive firefighting to predictive management-networks that self-monitor and self-heal reduce the volume of escalations that land on desks. Second, it compresses customer interactions from multi-stage funnels into single, intelligent conversations, which means operations and customer experience metrics become the same conversation. Third, success is measured by outcomes, not deployment scale. Banerjee pointed to operators cutting resolution times by 63% and lifting first-call resolution and NPS by 50% each-but only when AI was embedded across the entire customer journey, not dropped into isolated points. The operators that win will embed AI into everyday workflows, measure what actually improves, and keep human judgment on the decisions that matter.
Your membership also unlocks:
