Blue Planet’s configuration management platform is only the starting point. As telcos pursue autonomous networks, the challenge is not simply deploying AI agents, but giving them the data, context and governance needed to make trusted decisions.
In sum – what to know:
Beyond configuration – Blue Planet’s CCM platform addresses configuration drift and operational risk, but the wider ambition is to connect orchestration, assurance, analytics, and automation in a closed-loop architecture.
Context challenge – Telecom operators have no shortage of data, but fragmented systems and missing operational context have limited its value. Agentic AI can help, but only when grounded in reliable network intelligence.
Intelligence layer – Blue Planet says OSS will evolve from operational support systems into autonomous intelligence systems, where AI agents can plan, optimise and manage increasingly complex networks.
Picking up with Blue Planet again, a few weeks on: discussion about the Ciena-owned firm’s new configuration and change management (CCM) product has turned into a broader discussion about how telcos ultimately go about delivering holy-grail level-five (Level 5) autonomy. The immediate problem is familiar – just keeping complex multi-vendor networks stable as manual intervention, fragmented tooling, and operational silos create risk. The question is what happens next: how operators turn AI from an efficiency tool into a trusted operational layer for the all-critical network itself. The answer, says Blue Planet, is in the architecture around all the agentic shenanigans.
In the end, the connectivity discipline is about just that: connectivity – to connect orchestration, assurance, analytics, configuration, and the operational data that animates AI systems to make AI decisions. Configuration management is a foundation; autonomous intelligence is the ambition. Blue Planet wants to talk about the whole loop. Autonomous networks need intelligence across every silo: planning, provisioning, assurance, optimisation, and remediation – all working in concert. The point is not simply to automate configurations, but to understand why they are required, what effect they have elsewhere in the network, and whether they achieve the desired outcome.
Tracked auto-configuration provides context, but has to be mapped into closed-loop change management across the network, and across time. Blue Planet has the wraparound as well, says Gabriele Di Piazza, vice president of products and alliances, continuing the conversation from before. “We have an orchestration engine to make changes to the service, and an assurance system to understand the conditions and telemetry… We want to use configuration to enrich the fulfillment process, while maintaining visibility of configurations, and tracing changes as a symptom of outages, or whatever.” If performance degrades and parameters are reset, the changes have to go up the chain.
He says: “This is not just about managing a configuration, but about the attributes of service changes over time, and what happens three levels up – and then coordinating across multiple endpoints. That ability to track and audit the changes is something that has been missing in the industry. Often, the problem is not really day-zero when you make the configuration, but the lifecycle of it, and the drift – and how to manage all of it: comparing configurations, going back to compliance. Often when companies manage fleets of devices, a superuser logs into a router and makes a change, which creates lots of issues… You see alarms and faults, and don’t really know where to look.”
The approach reflects a wider shift in OSS architecture – and a shift that has also been discussed for decades, about how to normalize data across data silos and data lakes, and to leave it mostly where it is, and pull it from distributed infrastructure into connected cloud platforms as required. Telco networks have generated vast quantities of operational data for years, across inventory systems, assurance platforms, service records, support tickets, technical documents, vendor-specific tools. The challenge has never been to collect data, but to make sense of it. “The idea was to have a single data model, but that never happened – and probably never will,” he says.
“Everybody wants to differentiate.” What has changed, says Di Piazza, is not the nature of the problem but the ability to tackle it. Frontier AI models and agentic AI frameworks can sift structured and unstructured data from technical documents, engineering records, and service tickets at lightning pace. “You have different firepower,” he says. “The problem is more urgent because there’s more data, devices, sites, services – more velocity.” At the same time, he is careful not to make out like AI is a cure-all, off the bat. “Agents don’t solve everything; but they can really help to compare configurations, build playbooks, understand root causes. And adoption by telcos has really taken off.”
The other caveat is that AI agents live and die based on the quality of the data available to them – which brings us back to the data architecture, and the logic to leave most data where it lives, rather than to pull it into a single repository. Moving petabytes of live telco data into a centralized data lake is expensive and rapidly obsolete. “The latter,” responds Di Piazza, when asked if Blue Planet advocates unified platforms over unified databases. “You want that graph view, that digital twin view, where you can plan, test, and evolve.” But again, the trick is to preserve the decisions made against the data – the operational reasoning for the original diagnosis, adjustment, and resolution.
In a form that AI can use. Blue Planet’s vision is to build a decision framework alongside the network model itself, allowing future agents to compare new situations with previous outcomes, understand what actions were taken, and learn which interventions proved successful. “The agents are only as good as the data you give them,” repeats Di Piazza. “They are only as good as the context you build for them.” It is perhaps the clearest expression of where Blue Planet sees OSS heading. Its CCM platform, announced in June, is significant not just because it addresses one of the industry’s keenest headaches, but because it stops it from infecting the wider operational environment.
Configuration management might reduce outages and remove manual work, but it is a means to a larger end. Planning, inventory, orchestration, assurance, analytics, compliance, and configuration cease to be separate operational silos; they together become the knowledge base from which AI agents can make informed decisions.
For Blue Planet, that represents the next evolution of OSS itself. “We are in the automation business,” reflects Di Piazza. “Automation has gone from task-based automation, to automated configuration, and now to AI-driven automation. OSS will not just be an operational support system; it becomes an autonomous intelligence system.”
The future of autonomous networks is about giving AI enough context to make trustworthy decisions.
