Unryo Rebuilds 5G Signaling into Sequence Diagrams and Marks Failures

Unryo's AI agent analyzes SBI traffic for an AMF pod, plotting the procedures as a sequence diagram with the failed step in red, next to the health of the infrastructure beneath it.

Starting from a Kubernetes issue, Unryo pulls the corresponding traces and rebuilds each procedure as a sequence diagram, with timestamp and HTTP status on every exchange.

Service-based interface analysis that locates each failed 5G procedure down to the network function and pod

MONTREAL, QC, CANADA, September 29, 2026 /EINPresswire.com/ -- Unryo has released agentic AI analysis of 5G core signaling, which reconstructs probe traces into per-procedure sequence diagrams and identifies where a procedure failed or slowed down.

Control plane troubleshooting in a 5G core is still largely manual. An engineer pulls a trace, works out which exchange failed or ran slow, and then checks the network function and the infrastructure beneath it in separate tools. The work is slow, it is repeated for every incident, and the result lives in one person's notes.

Unryo's AI agents do that work on request. An engineer asks in plain language for the signaling of a given subscriber identity (IMSI) over a chosen time window, at any time and with or without an open incident. The agents retrieve the matching traces and assemble them per procedure, including registration, PDU session establishment and handover, into a sequence diagram that shows every request and response between network functions in chronological order.

A failed step is marked with its response code and the network function instance that returned the error. The agents also flag missing messages, such as a lost packet, and use 5G protocol context to judge whether each response time is acceptable.

When the agents find a problem, they use the platform's topology to identify the infrastructure element serving that network function, down to the pod, and check whether that element is saturated, which would make it a probable cause.

Investigations start from the failing step and its probable cause, not from a raw trace, so engineers spend their time on the fault instead of on piecing the investigation together across tools. Every query the agents run is shown alongside the result, so the evidence behind each diagram can be checked rather than resting on one person's notes.

About Unryo

Unryo is a full-stack observability and agentic AI platform for network operators. Its Topology Data Fabric discovers and maps multi-layer dependencies across physical, logical, and application infrastructure, giving AI agents the context to find root cause and drive resolution.
Available now, deployable on-premises or in an operator's private cloud.
www.unryo.com

Virginie Porter
Unryo Inc.
info@unryo.com