Can Your Network Handle AI Agents? (EP 1089)

About this episode

Before you buy another AI tool or help a client deploy agents, take a hard look at the environment supporting them. Recording from Greenville, South Carolina, I talk with Joseph Vlajcic about what happens when businesses chase individual AI use cases without considering how those systems will work together. We discuss why disappointing returns can come from more than choosing the wrong tool. Sometimes the surrounding infrastructure needs attention first. Joseph walks through AI running in the data center, public cloud, and at the edge, with the network connecting all three. That brings up questions MSPs need to ask about traffic, access, security, and visibility. I also raise something we haven’t spent enough time discussing: we’re watching AI-driven threats coming in, but what about the activity our own AI tools generate going out? We wrap up with the business conversations that can be harder than the technical work. How do you handle clients who want AI but resist your recommendations? What belongs in the discussion around service agreements? And what should you never promise about an AI future nobody can fully predict? Joseph brings it back to segmentation, identity, patching,