The team builds b.lo, a network performance platform for operators, across access, transport and core. It collects every source, from every vendor, into one data model, and splits into independent modules that work together: three business modules your engineers use every day, four configuration modules to build the tools they need, and one to administer the whole. On top, query lets them question all of it in plain language. Underneath, an execution harness rules on every action that goes back to the network. Every KPI, alerter, process and report sits in a shared library, so the best expertise spreads across the team.

How agentic AI changed the way we approach network monitoring.

For years we have monitored, analysed and optimised operator networks, radio, core and transport, across every vendor. Machine learning has been part of our culture. It sorts, clusters and filters the lists that our customers' engineers work through every day. The lists themselves, even well filtered, still had to be worked one line at a time, by people.

Agentic AI changed everything in 2025. That tedious work can now be handed to agents that read the data, walk the knowledge tree and come back with a proposal and its evidence. For a team that had spent years building those lists, it was a genuine break: the service stops being a screen to read and becomes a decision to take.

Everything on this site follows from that, and from one conviction that came with it. An agent that acts on a critical network must be held by something stronger than a prompt.

Headquartered in Toulouse – France