An article by Joe Cumello, Senior Vice President & General Manager at Blue Planet.

Our industry is at an inflection point. As we design the OSS of the future around Agentic AI, we can either follow open autonomous networks principles or we can recreate the vendor-specific silos of the past. On the back of deep customer conversations and extensive development work, Blue Planet is introducing our Agentic AI Framework, in the hope that it helps our CSP customers architect next-generation OSS with the agility, efficiency, and openness required to create great customer outcomes and measurable operational benefits. Joe Cumello, SVP & GM of Blue Planet, explains.

I think we can all agree that AI content has reached unmanageable levels of hyperbole. So, it’s with a bit of trepidation that I come to you with another AI-focused discussion.

The struggle we’re all having is getting to the bottom of whether — or maybe even how — agentic AI can transform communication service providers’ (CSPs) network operations. We’re all operating under the premise that AI can and will make a difference, but how do we provide the industry with an answer and avoid the mistakes of the past? Is there anyone sitting back thinking about how to implement a solution that goes beyond untold numbers of vendors all offering their own agentic AI answers in a point solution context? Aren’t we at risk of repeating the ‘sins of the past’ of a fragmented legacy network?

Let’s assume you’re not interested in that happening.

This is why we are introducing a different approach based on conversations with many of you over the last year: Blue Planet’s Agentic AI Framework. The idea is to deliver an agentic AI framework, purpose-built for telecom networks that delivers AI agents that act with intent, apply context, and take coordinated actions across your entire network. But they do this on top of an underlying and converged data model that is clean and organized. As we often say, ‘You Can’t AI What You Can’t See’. Applying AI agents against bad data doesn’t serve anyone.

So we’re pretty convinced this isn’t an AI strategy duct-taped onto legacy OSS. It’s an approach and a platform that brings together our decades of telco expertise, a federated data foundation, and advanced AI capabilities to unlock the full value of autonomous networking.

Intrigued? Let’s explore what we’re proposing:

Figure 1. Blue Planet’s Agentic AI Framework

Why Blue Planet’s Agentic AI stands out

Access to the right data across the entire CSP network

AI agents are only as effective as the data they consume. To deliver real impact, CSPs need access to the right data — not just more of it — with end-to-end visibility across their network. Blue Planet has a decade of expertise consolidating, federating, and simplifying access to inventory data, providing a single source of truth on which CSPs can orchestrate, automate, and assure their networks. This structured and stateful data fabric – across multiple domains and vendors – allows AI agents to act and take appropriate actions in a timely manner.

Enabling AI agent coordination – not siloed automation

Truly achieving OSS modernization doesn’t lie in isolated automation. It requires full network visibility and coordinated agents that collaborate across the OSS domains to drive real transformation. While many vendors offer AI agents to streamline tasks in a particular area or function, these siloed solutions offer short-term efficiencies, but the broader structural limitations of legacy OSS environments remain.

At Blue Planet, we believe in taking an open and holistic OSS agentic approach — normalizing and unifying cross-domain, multi-vendor data with embedded OSS knowledge. Within our Agentic AI Framework, agents can operate with awareness of the broader network context and business priorities, executing goal-driven tasks that are orchestrated, efficient, and aligned. Leveraging our data model, agents can operate from a single source of truth. This level of coordination is only achievable with end-to-end network visibility and clean data that breaks down data silos — providing the strategic foundation for a truly holistic, next-generation AI-driven OSS.

Accelerate time to OSS value

You could build your own agentic AI capabilities – there are lots of general-purpose AI tools available after all. But doing this would require knowledge of exactly which OSS datasets are needed, in which format, and would require months of data preparation, model training, and integration. How do you anchor the agentic capabilities in your current processes? If you build on an external platform, you’ll have to work with your vendors to integrate the outcome the agents drive into their applications or create another tool for your users.

Blue Planet delivers a direct path to operational value through our holistic approach. With Blue Planet AI Studio, we offer an integrated development and execution environment within our OSS platform, and because of this, you can build agents to perform specific operational tasks, using the right datasets and processes, and immediately deploy them within inventory, orchestration, and assurance applications.

For an even faster time-to-deployment, we offer pre-packaged AI agents designed to solve common operational problems and OSS-specific tooling all within a low-code/no-code environment. You’ll still have the flexibility to bring your own models and integrate the LLM of your choice to meet the compliance, security, and deployment requirements of your organization.

The figure below highlights some of the pre-packaged AI use cases, CSPs will be able to use to transform their network operations.

Figure 2. Agentic AI Use Cases

We’re not distracted by generic AI experiments or chasing LLM hype; we’re focused on solving real challenges in telecom operations, where time-to-impact and business outcomes matter most.