SensyAI: AI-Powered Troubleshooting for FTTH Networks

One search. The whole path. A root-cause verdict in seconds.

SensyAI is the AI troubleshooting layer of NetSense. Give it any customer identifier and it walks the entire service path, from upstream and core through the BNG, distribution, and OLT down to the ONU and the WiFi in the living room, then tells your team exactly what is wrong and on whose side of the network it is.

From raw device data to resolution

Your network already produces the data. SensyAI turns it into answers, for your NOC, your support line, and soon your customers directly.

SwitchesRoutersBNGsOLTs & ONUs
NetSense NMSSNMP · Syslog · Streaming Telemetry
Edge RoutersCore Routers
NetSense FlowNetFlow · IPFIX · sFlow
CPE & WiFiONTs
NetSense PulseTR-069 / ACS · WiFi telemetry
SensyAIwalks the whole path: Upstream → Core → BNG → Distribution → OLT → ONU → WiFi
Support call
Agent searches any IDPPPoE user, serial, or MAC
Full-path diagnosis + AI verdictISP-side vs in-home, in seconds
Fixed on the first callno blind ONU reboots
Network fault
SensyAI isolates the faultOLT, fiber, uplink, or switch
Ticket with root cause attachednot just "customer says slow"
Field engineer sent to the right spotone truck roll, not three
Customer chatbotSOON
Customer asks the chatbot"why is my internet slow?"
Same brain answersin-home WiFi vs network, instantly
Self-service fixno call, no queue
Every check SensyAI runs uses data NetSense already collects. No new agents, no new collectors.

How a diagnosis works

SensyAI runs the same path walk your best engineer would, in seconds instead of an afternoon.

1

Search anything

PPPoE username, ONU or CPE serial, or a MAC address. SensyAI resolves the customer even when the identifier is partial or vendor-mangled.

2

Core and access switches

Status, uptime, alarms, and port errors or discards on the switches feeding the OLT, so an uplink problem never masquerades as a customer problem.

3

OLT health

Device status, recent reboots, active alarms, power supplies, and the online ratio of every ONU on the port.

4

ONU and fiber

Optical receive and transmit power against thresholds, temperature, QoE score, and 24 hours of down events, with dying gasp (power) separated from optical faults (fiber).

5

CPE and WiFi

WAN status, CPE load, and per-device WiFi signal strength with band, straight from Pulse over TR-069.

6

AI verdict

A plain-language conclusion in three parts: what the customer is experiencing, what is wrong on the ISP side, and what is wrong in the home.

What SensyAI brings to your support flow

Search by anything

PPPoE username, serial, or MAC. Smart resolution finds the customer even from a partial or derived identifier.

Full-path correlation

One diagnosis covers the path from upstream and core through BNG, distribution, and OLT to the ONU, CPE, and WiFi. No swivel-chairing between tools.

Dying gasp vs fiber fault

Power outages at the customer premises are automatically separated from real optical faults, so nobody chases a fiber cut that is actually an unplugged ONU.

ISP-side vs in-home verdict

The AI conclusion tells the agent which side of the demarcation point the problem is on, and what to do about it.

Live streaming diagnosis

Results stream in phase by phase as SensyAI walks the path, with a live network diagram the agent can follow.

Built on your NetSense data

No new agents or collectors. SensyAI reads the telemetry NetSense NMS and Pulse already gather from your network.

SensyAI FAQ

Common questions about the AI layer of NetSense.

Written by Plamen Haralambiev, Network Engineer and ManagerLast updated: July 7, 2026

See a Live Diagnosis on Real Data

Book a demo and watch SensyAI walk a real customer path, from upstream to WiFi, and call the root cause in seconds.