Network Copilot details Aviz Service Node analytics integration with ELK chat
Network Copilot 1.7.0 adds an integration for Aviz Service Node (ASN) subscriber and application analytics by using the existing ELK data connector, enabling chat-based queries over Elasticsearch session data without custom dashboards.
Research Overview
The blog explains how telco operational analytics typically reside in dashboards and require either custom queries or knowledge of specific data paths. It describes a workflow in which ASN analytics and session metadata are exported to an Elasticsearch stack and then queried through Network Copilot chat.
The approach is positioned as using the same ELK connector already present in the Network Copilot environment. After connecting to the Elasticsearch instance and specifying the ASN index pattern, operators can query session-level telemetry via natural language.
Key Findings
With Network Copilot 1.7.0, ASN data becomes queryable in chat at the session level by aggregating fields from a session-level Elasticsearch index. The blog indicates the resulting dataset can include tens of millions of session documents and, in one example, about 11.1 TB of traffic for the bandwidth query.
The blog describes that operators can ask for subscriber and radio/mobility analytics, application analytics, and performance-oriented results such as latency, packet loss, and jitter. It also states that answers can be returned as text, tables, or interactive charts like pie and bar charts, depending on the query.
Technical Breakdown
The integration relies on ASN continuously exporting application and subscriber metadata to an Elasticsearch stack. Network Copilot connects to the same ELK stack using the ELK data connector, making the ASN session index available for natural language querying in chat.
Data is stored in a session-level index, such as a user-session-data index, with documents that contain ASN metadata including destination, transport, and KPI fields. The blog describes these per-session fields as enabling detailed query responses at session granularity, including ranked results.
Operational Impact
The blog provides examples of operator questions that start at broad summaries, such as total bandwidth consumed, and then narrow into subscriber-specific views. It lists subscriber queries including total subscriber count, top consumers, bandwidth per subscriber over the last 24 hours, RAT-type distribution, handover counts, and tower density.
For application analytics, the blog describes queries about top bandwidth-consuming applications and breakdowns by application category, such as Video, Gaming, Audio, and VoIP. It also describes engagement and performance views, including highest-latency application types and ranking the worst-performing sessions by a combined KPI score, returning a sortable table with fields such as source and destination IPs, ports, protocol, application name, maximum packet loss, and jitter.
Leadership Perspective
The post frames the change as reducing reliance on manual dashboard navigation and specialized query tooling by shifting interaction to conversational queries. It asserts that no additional data modeling or dashboard creation is needed beyond pointing the ELK connector to the Elasticsearch instance and the relevant ASN index.
For enterprise IT and SOC-style operational workflows, the blog highlights that session-level telemetry can support both exploratory questions and targeted troubleshooting outputs in the form of ranked and exportable tables. It also notes that operator interaction occurs through chat once the ELK connector is configured under Data Connectors.
Network Copilot Release 1.7.0 integrates Aviz Service Node analytics into chat-based querying by connecting to an Elasticsearch stack through the ELK connector and using ASN-published session-level telemetry. This fact-based blog signals brief reporting on how subscriber, radio/mobility, and application performance data can be queried in natural language for enterprise decision-makers, including CISOs, CIOs, and SOC managers.
Source: aviznetworks.com, by Vishakh Dhar.