BBL-NET
Support automation on a real ISP
An AI support system wired into UISP and Odoo that identifies the subscriber, diagnoses the link against live network data, and resolves the routine load in a median of 7 seconds, 24/7.
- 7s median reply
- 24/7 in production
- 7,000+ messages handled
- UISP + Odoo integrated
Context
BBL-NET, a wireless ISP of ~800 active clients, faced the same flood every small ISP does: repetitive support around the clock. Hiring doesn’t scale; generic assistants don’t know the network.
What we built
A support system that identifies the subscriber against the CRM, runs a real diagnosis against UISP NMS (outages, CPE and parent-AP state, signal and traffic vs. history) before responding, resolves billing/debt/suspension/install questions, and escalates cleanly with a UISP ticket when needed — never inventing data.
Results
Median reply time of 7 seconds, running 24/7 including overnight. Over 7,000 messages handled in recent months. The large majority of contacts are resolved without a supervisor stepping in. In production, integrated with UISP (CRM + NMS) and Odoo.