Building AI agents for ISP networks sounds straightforward until you account for the reality of what those networks actually look like. Preseem co-founders Scot Loach (CTO) and Dan Siemon (CEO) dig into the architectural challenge that most AI discussions in the ISP space quietly sidestep: the extraordinary fragmentation of equipment, telemetry, and data across a typical operator’s infrastructure.
A single ISP might run access points, point-to-point radios, OLTs, and home routers from four or five different vendors, each generating telemetry in different formats, using different terminology for the same concepts, and exposing data through incompatible interfaces. Acquisitions make it worse.
This session goes below the use-case layer to examine what’s actually required to build AI agents that work reliably across this complexity, and why the normalization and context layer sitting between raw network data and AI models is the piece most operators are missing.
STOP MANAGING COMPLEXITY
START RUNNING YOUR BUSINESS
Every week without visibility is another week of trucks that didn't need to roll and subscribers who left without telling you why.
Let's start with your network, your vendors, your subscribers.
© Aterlo Networks 2026 | aterlo.com
