Sub-Saharan Africa telco~1,000 sites · 2.3M subscribersPowered by NexusAI

Passive Network — Detection & Revenue Recovery

Ontology × Network topology × AI = Proactive detection

Projected ROI · Proactive detection, revenue protection, MTTR reduction and dispatch optimisation
40%MTTR reduction
$3.15M–$4.75MTotal annual impact
{ How Passive Network works }

From raw signals to decisions

Data sources
RAN performance filesEricsson 2G cell metrics
OSS alarm streamsReal-time fault events
ENIQ data warehouse85,000+ tables, 14-day rolling window
BSS subscriber recordsRevenue & usage data

Passive Network AI engine

Ontology-driven reasoning over Cell → Neighbour → Transmission → BSC chains

Ontology mappingAnomaly detectionNeighbour analysisRoot-cause MLRevenue meteringAuto-remediation
Insights delivered
Detection alertsConfidence score & reason
Revenue impact$ per hour per incident
Root-cause reportsFailure category & evidence
Action plansTiered remediation steps
70%+passive cells detected proactively
40%MTTR reduction achieved
4–6 hrsdetection time eliminated
30–40%unnecessary dispatches avoided
90 daysfull deployment, no lock-in
{ The challenge }

What stood in the way

The situation before NexusAI — in the operator’s own numbers where they exist.

  1. 014–6 hrs

    Reactive Detection

    Passive network cells found only after customer complaints pile up — 4-6 hours after degradation begins.

  2. 02—

    Data Fragmentation

    Intelligence scattered across Ericsson RAN, OSS, ENIQ, AUTIN and BSS — no unified view exists.

  3. 03—

    No Revenue Awareness

    Faults have no link to financial impact — NOC works by queue order, not business value.

  4. 04—

    Tribal Knowledge Detection

    Engineers manually query for 100% availability + zero traffic and review trends by eye.

  5. 0530–40%

    Avoidable Field Dispatches

    30-40% of truck rolls are unnecessary — root cause could be verified remotely first.

  6. 068–24 hrs

    Slow Resolution Cycles

    8-24 hour MTTR depending on complexity, with no learning loop to accelerate future fixes.

{ The solution }

What NexusAI delivered

An ontology-first AI layer deployed in 90 days on the operator's live Ericsson 2G network — detecting passive cells within minutes, calculating real-time revenue impact and auto-remediating where safe. No systems replaced.

  1. 01

    Ontology-Driven Detection

    Continuously monitors all 1,000 cells via Cell→Neighbor→Transmission→BSC relationship chains. Flags passive conditions within minutes.

  2. 02

    Neighbor Overflow Analysis

    Queries adjacent cells for traffic spikes confirming load redistribution — correlates active alarms on transmission links.

  3. 03

    Root-Cause Identification

    Traverses ontology to identify failure category (transmission link failure vs. software fault) with quantified confidence level.

  4. 04

    Real-Time Revenue Metering

    Calculates affected subscribers, revenue loss per hour, and projected total loss if unresolved for historical average MTTR.

  5. 05

    Automated Remediation

    Executes soft-resets where ontology confirms safety. Escalates to L2 or field dispatch only when physical fault evidence requires it.

  6. 06

    NOC Dashboards & Knowledge Transfer

    Real-time cell health map, active alert queue, ontology graph visualization, and operational metrics with AI reasoning chain.

{ Results delivered }

The outcome

70%+Proactive detection

Passive cells detected before any customer complaint — versus 0% proactive detection before.

40%MTTR reduction

AI-guided investigation replaces hours of manual data-gathering across siloed systems.

50%Investigation time cut

Automated root-cause analysis in seconds via ontology traversal.

$2.6M–$3.9MRevenue protected

Annual revenue at risk from passive cells now monitored and defended in real time.

30–40%Dispatch reduction

Remote verification confirms root cause before sending technicians — $150K–$250K saved a year.

90 daysFull production

Entire capability live in production within the engagement window — no vendor lock-in.

← Previous case studyNOCSenseNetwork operations intelligence Next case study →SupplyChainSupply chain intelligence
Get started

Get results like Passive Network

Turn enterprise systems into an intelligent operating layer that understands your business and executes operations autonomously.

Request a demoTalk to sales
Live in production within 90 daysNo systems replacedCloud, on-premise or hybrid