Effective Network troubleshooting is one of the top, if not the number one, priorities for almost every network operations team. Having the tools and the skills to identify faster the root cause and impact of network issues has a direct connection with customer satisfaction.
Automating troubleshooting has always been difficult because this task is highly knowledge and decision centric. Other network activities such as doing software upgrades, for example, can more easily be automated. Knowledge and decision making are inherently human, so the mechanisms to help us automate these tasks are not common.
AI Agents seem like an interesting building block to automate Network troubleshooting. Because they are software based, they bring new angles we could not think of before. What if we have 100 Agents in parallel troubleshooting a problem? or 1000? This is just an examples of the possibilities that software brings.
We are definitely exploring using Agentic AI for network troubleshooting. It is promising.
I wonder what are your thoughts about using AI Agents for network troubleshooting? Any experience so far?
