Automate network operations smarter and faster with agentic AI

 AI is at the forefront of the technology market sector these days. Research & Development companies are producing more powerful GenAI models, and new companies are appearing on the scene (witness DeepSeek). Global content providers and service providers aren’t standing still either. They are rearchitecting and scaling their networks to ensure they can accommodate the requirements that AI demands:

• Hyper-connected hyper-powerful GPU clusters for the training of Large Language Models (LLMs)
• High-bandwidth connectivity for real-time inferencing to respond to user’s queries

In addition, global content and service providers are exploring how

 AI can be leveraged to improve and accelerate their network operations – so called AIOps. In a previous blog, I discussed some high-runner use cases with Ciena experts. In general, AI, Machine Learning (ML), and GenAI can be applied to three key areas to empower network operations teams:

• Assurance
• Optimization
• Automation

In each of these areas, today, germany telegram data a network operator not only needs to know what they want to do specifically, but they also need to know which tools to use and how to use them. And what if they don’t even know that the tools exist? Enter the Navigator AI Assistant – the AI-powered natural language interface to our Navigator Network Control Suite (Navigator NCS).  Now, you only need to know your intent, ask the question(s), and then AI Assistant figures out which tools and capabilities to use to answer your question. This is analogous to the sea change that occurred when we went from manually leafing through encyclopedias to typing in questions into a web browser, which automatically returned the best answer it could find from the vast Internet knowledge base!

In the context of network assurance, 

Optimization, and automation, there is, of course, the underlying requirement that correct tools and data are available; otherwise, the generated response has little merit. That’s why it’s imperative to have an appreciation of the intelligence that is built into a network control system. In particular, it’s essential to start from a position of expertise in advanced analytics.

In today’s Navigator NCS, there is a myriad of AI-driven capabilities for network utilization trends, configuration audits, problem analysis, pinpointing fiber faults, service health insights, and more. And all of these are presented within an intuitive GUI interface or are programmatically accessible via APIs. The AIOps secret sauce is in tying together the right datasets with the right tools in the right sequence, at the right time, in order to address the operator’s intent rapidly.

How is this done?

 The key is the use of multiple AI agents, with an overarching orchestration agent, to deliver a dynamic workflow, as depicted in Figure 1 below. Each agent is responsible for its own scope, drawing on source stored data (either in databases or embedded within network elements themselves) to perform reasoning, make a decision, and take action. This is known as agentic AI. The human operator participates in the dynamic workflow by verifying the results and initiating multiple loops as required, leveraging the easy-to-use natural language interface.