Enabling network issue detection on scale

Overview

In 2025 Nokia introduced ANF to advance autonomous network management. I worked end to end to shape workflows that helped engineers find the source of service issues faster, reducing the time customers spent affected.

In 2025, Nokia introduced Autonomous Network Fabric as part of its shift toward AI-enabled networks. Within this platform, Subscriber observability focused on how telecom operators understand and manage service quality at the subscriber level.

Impact

Time from issue detection to resolution fell by 25%

Employer

Nokia

Team

Analytics

Tools

Problem

The network teams responsible for keeping mobile service reliable had no clear path from a subscriber complaint to the underlying cause. They had to piece together clues across multiple tools, making every investigation slower, more uncertain, and harder to act on with confidence.

Approach

Vodafone came to us looking for better visibility into degraded subscriber service. As we mapped how network teams actually investigated these issues, we saw how much work happened outside any single tool.

Engineers were moving between dashboards, checking 25+ KPIs, reviewing sessions and network conditions, and taking manual notes to connect the evidence.

Vodafone came to us looking for better visibility into degraded subscriber service. As we mapped how network teams actually investigated these issues, we saw how much work happened outside any single tool.

Engineers were moving between dashboards, checking 25+ KPIs, reviewing sessions and network conditions, and taking manual notes to connect the evidence.

That changed the problem for us. Bringing everything into one place wasn’t enough — the product had to help engineers understand what mattered first, what to look at next, and how to narrow the investigation without exposing all of the complexity at once.

I explored different starting points across subscribers, poorly performing cells, and geographic context to find the structure that best supported how engineers actually diagnosed issues.

That changed the problem for us. Bringing everything into one place wasn’t enough — the product had to help engineers understand what mattered first, what to look at next, and how to narrow the investigation without exposing all of the complexity at once.

I explored different starting points across subscribers, poorly performing cells, and geographic context to find the structure that best supported how engineers actually diagnosed issues.

Solution

Tracing service issues from affected subscribers to network signals

We designed Observability experience as one continuous investigation workflow with a clear drill down appraoch — Define the scope, identify affected subscribers, locate where and when the issue occurred, review the relevant sessions, and compare the network signals behind them.

Idea 1

Map-first investigation

We explored starting from problematic areas on the map and correlating KPIs with subscriber experience.

It provided strong network context, but made it harder to see who was affected, so the map became supporting evidence rather than the entry point.

Idea 2

KPI-first appraoch

We explored letting engineers select a cell, compare subscriber KPIs, and then drill into individual sessions.

We didn't persue this as this still required them to manually find patterns and determine who was affected.

Impact

  • Time from issue detection to resolution fell by 25%.

  • Contribution to design system & design patterns for the portal.

  • Setting up AI assisted workflow at Nokia.