Unryo turns 5G signaling traces into sequence diagrams with failure detection

Unryo has launched agentic AI analysis for 5G core signaling that rebuilds probe traces into procedure-level sequence diagrams and flags failures, slow responses and likely infrastructure causes. The tool is aimed at speeding up control-plane troubleshooting for network operators by linking each issue to the specific network function and pod involved.

Why it matters: - 5G core troubleshooting often starts with a raw trace and ends with manual work across multiple tools. - Unryo’s approach aims to shorten that loop by identifying the failing step, the likely cause and the infrastructure layer behind it. - The result is faster control-plane incident handling and less reliance on individual engineers’ notes.

What happened: - Unryo released agentic AI analysis for 5G core signaling. - The system reconstructs probe traces into per-procedure sequence diagrams. - The analysis works on request from an engineer using plain language. - Engineers can ask for signaling tied to a specific subscriber identity, or IMSI, over a chosen time window. - The workflow is available with or without an open incident.

The details: - The agents retrieve matching traces and organize them by procedure. - Supported procedures include registration, PDU session establishment and handover. - The output shows each request and response between network functions in chronological order. - Failed steps are marked with response codes and the network function instance that returned the error. - The system also flags missing messages, including lost packets. - 5G protocol context is used to judge whether a 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. - The analysis can trace that infrastructure element down to the pod level. - The agents check whether the element is saturated when saturation could explain the failure. - Every query the agents run is shown with the result, so engineers can verify the evidence behind each diagram.

Between the lines: - The product shifts troubleshooting from trace interpretation to fault investigation. - That matters because control-plane issues often span network functions and the underlying infrastructure, not just one log or one node. - The platform’s topology context is what turns a signaling failure into a likely root-cause path.

What’s next: - Unryo says the capability is available now. - The platform can be deployed on-premises or in an operator’s private cloud. - Unryo positions the system as part of its broader observability and agentic AI stack for network operators. - The company’s Topology Data Fabric is designed to map dependencies across physical, logical and application infrastructure so AI agents can support root-cause analysis and resolution.

The bottom line: - Unryo is trying to make 5G signaling troubleshooting faster, more explainable and more actionable by turning traces into procedure-level diagrams with built-in failure analysis.