One of the clichés in sports is that the best teams do the basics well. Only once you master these basics will you be able to focus on your results. Doing the basics well and doing simple things well are not the same thing. The basics are not easy. Mastering the basics requires discipline, patience, and a deep understanding of the fundamentals of your role.
Currently, the industry conversation is full of promises regarding the integration and deployment of AI technologies in network operations, including AI-native networks, intelligent OSS, agent automation, and autonomous operations. Driven by investor pressure, the “race” for AI relevance, and strategic priorities such as technology transformation, the focus on AI-native network outcomes is understandable.
Joe Kumelois SVP and GM of Blue Planet, the communications software business within Ciena. Last year, Blue Planet’s revenue increased 49% year-over-year to $115 million, driven by sales of inventory, orchestration, and assurance software that supports the transition to an automated network. As our article “The Battle Over AI and Data Management” reported last year, the company also introduced Agent AI Studio in 2025.has already won in that area.
Cumello believes that AI-driven autonomous networks represent a shift in industry paradigm alongside previous shifts such as SDN, virtualization, and cloud. But delivering on that promise requires a focus on the basics, he warns..
“Success doesn’t come from marketing slogans or ambitious announcements. It comes from the foundation of trusted data, secure architecture, and a platform that gives operators control.”
The first of these fundamentals is producing and accessing clean, context-aware data. Cumello says that before operators deploy advanced AI agents and autonomous workflows, they need to address a long-standing issue: data quality.
“AI is only as good as the data it works with,” says Kumelo. “You can announce all the AI platforms you want, but none of them will work if you don’t have clean, context-sensitive data in the domain you operate in.”
As Cumello explains, for years, many carriers have centralized their information in modern data lakes. It’s a useful first step. However, simply aggregating data is not enough. Communication environments are highly interconnected, and AI systems need to understand their relationships.
There it is Context-aware data comes in. This is important because traditional systems often contain inaccuracies. Old inventory platforms may still list closed sites, obsolete equipment, or “ghost” infrastructure. When AI operates on bad information, it can make costly mistakes, such as dispatching truck rolls to the wrong location or misallocating resources.
But it’s not just about knowing what assets exist. Context-aware data means understanding how services are connected between optical, IP, and mobile layers, how SLAs are linked to infrastructure, and how systems interact operationally.
secure the foundation
Security is another foundational piece, a foundational piece, that cannot be put on the back burner, especially in an autonomous environment.
As networks become more self-operating, their risk profile changes as well. AI agents gain the ability to make operational decisions automatically. This can create new efficiencies as well as potential for misuse.
“If an attacker injects rogue agents or corrupts the data that feeds those agents, it could lead to a network outage or a ransom-type scenario,” Kumelo said.
The goal is not to slow down innovation, but to design security into the architecture from the beginning and have clear control over who can deploy agents, how they are validated, and the guardrails that control agent actions.
There’s often more talk than action when it comes to security-by-design approaches, and Cumello acknowledges that the industry hasn’t fully resolved the issue yet, but the right action is to be proactive rather than add on later.
Democratize AI and don’t recreate old models
The third necessary change that Kumero describes is what he calls the democratization of AI.
Historically, OSS/BSS environments have relied heavily on professional services and custom development. Vendors have built complex systems that can only be modified by themselves or by expensive integrators. Change requests have become a source of revenue.
However, there is a growing desire for operators not to use that model.
“They don’t want to be dependent on suppliers every time they make a change,” he says. “They want to democratize decision-making and development.”
This means providing internal teams with the tools themselves to build and modify automation. According to Cumello, Blue Planet’s unique approach has always been to productize OSS capabilities and layer AI tools on top, such as low-code and no-code studios. Customers can access clean data, select or build agents, and design automation without ongoing vendor intervention.
“If the customer wants us to do the job, we’ll do it,” he added. “But the key is to empower them, not lock them up.” In this model, the vendor provides the platform and guardrails. Operators and integrators build intelligence.
financial pressure
As appealing as this technology is, the real motivation for AI transformation is not innovation for its own sake. It’s an economical thing. Telecom operators are facing slowing service revenue growth and rising operating costs. Investors and boards of directors are looking for higher profit margins. Autonomous networks offer a path to both faster service delivery and significant cost savings.
“The discussion isn’t about deploying AI because it’s trendy. It’s about cutting hundreds of millions or even billions of dollars from operations while increasing agility.”
This relationship between AI and measurable financial outcomes has fundamentally changed the customer conversation, says Kumelo. A year ago, many projects focused on replacing aging inventory and warranty systems. Operations managers now have bigger questions about redesigning operations, automating end-to-end processes, and completely eliminating traditional dependencies for large-scale cost savings.
How is the industry moving forward?
Despite the constant drumbeat surrounding AI, Kumelo said its adoption in the telecom sector is a reality. Telcos are not attempting large-scale transformation overnight. Instead, we work on a use case-by-use case to clean data, automate manual processes by agents, measure savings, and scale from there. This approach is deliberate rather than experimental, and focuses on real-world operational improvements rather than proofs of concept.
“It’s gradual,” Kumelo said. “We make sure our data is clean, we make sure we’re getting the benefits we expected, and then we grow.”
He added that Blue Planet already has these features in place for paying customers, and more are on the way. This approach is deliberate rather than experimental, and focuses on real-world operational improvements rather than proofs of concept.
We expect to hear more about this at MWC Barcelona and beyond.
