How one of India's largest neutral telecom infrastructure providers replaced a patchwork of siloed element managers with the NetSingularity OSS, unifying fault, performance, inventory, topology and service fulfilment across IP-MPLS, DWDM and legacy layers.
The Customer
The customer is one of India's largest neutral telecom infrastructure providers: a national-scale public-sector operator running a pan-India optical backbone. It delivers connectivity and managed services to government, enterprise and retail customers nationwide. Behind that reach sits an enormously diverse network: IP-MPLS routing, DWDM optical transport and a large installed base of legacy switching, built over years from more than a dozen equipment vendors. Keeping it running meant operating each vendor's element manager on its own island, with no unified view of the whole.
The Challenges
Every vendor came with its own management system. Operators were stitching the network together by hand: slow, reactive, and blind to the end-to-end picture.
Faults, performance, inventory and topology lived in separate vendor tools, with no single pane of glass across the network.
Tens of thousands of raw, uncorrelated alarms buried the real problems and overwhelmed the NOC.
Manual triage, manual ticketing and slow root-cause analysis meant longer outages and higher risk.
Services were turned up device by device over CLI, which was error-prone, inconsistent and hard to scale.
No unified, real-time record of physical and logical resources or the service topology built on them.
Legacy tooling couldn't keep pace with a growing, multi-technology, carrier-grade network.
The Solution
A single, cloud-native OSS that ingests from any vendor and delivers unified fault, performance, inventory, topology and catalog-driven service fulfilment across all three network layers.
Architecture
Built carrier-grade, open by design: Kubernetes-orchestrated microservices, a Kafka event backbone, and a multi-model data layer (relational, graph and time-series) engineered for 1M+ events per day and 54M performance counters per hour. Open, standards-based interfaces integrate cleanly with the operator's ERP (Oracle) and HPSM ticketing, so the OSS fits the existing operations landscape rather than replacing it.
Protocol-agnostic ingestion, inventory-enriched alarms, and multi-algorithm correlation (topology, parent-child, mass-outage and intermittent), with deduplication and automatic ticketing into HPSM.
Multi-vendor counter ingestion, a vendor-neutral KPI builder, node/time/hierarchical aggregation, dashboards, threshold alarms and on-the-fly analytics.
Automated multi-vendor discovery with OSPF/BGP/LLDP link mapping into a graph-based inventory, kept current by near-real-time reconciliation.
Unified logical and geographic views with end-to-end service visualization and drill-down from map to device to alarm.
Catalog-driven, workflow-automated provisioning of L2VPN, L3VPN, VPLS, VLAN and Internet, with pre-checks, validation and automatic rollback.
Self-service portal for service status, SLA visibility and ticket raising, extending OSS value to end customers.
One Consolidated View
The platform's core value: every layer, every vendor and every service the network runs on, normalized into one common model and one operational picture, instead of a dozen disconnected tools.
Domains
Vendors
Technologies & Services
netSingularity OSS: a single pane of glass
Common schema and vendor-neutral KPIs · unified inventory, topology, faults and performance · one console for the entire network
Order to Activation
Service requests arrive from the customer self-service portal or northbound from the ERP (Oracle), and the LCM Orchestrator turns each one into a repeatable, automated flow: feasibility checks, activation across multi-vendor devices, and full lifecycle control from request to ready.
Order raised via the customer portal or northbound from the ERP (Oracle).
A reusable service template and provisioning workflow are selected for the order.
Interface, neighbor, VLAN, MPLS and VRF feasibility validated before anything is pushed.
Configuration is pushed and activated across multi-vendor devices.
Real-time status, with automatic rollback or retry on failure.
The change: service turn-up moves from manual, device-by-device CLI to consistent, catalog-driven automation, with faster activation, fewer errors and full traceability across vendors and locations, and rollback to de-risk every change.
Results in the Field
Three views carry the day-to-day work on the live network.
A single monitoring screen tracks tens of thousands of alarms across routers, DWDM, ILA and OADM nodes, with live open/closed state, location and alarm code, so the NOC works from one correlated view instead of many.
Vendor-neutral KPIs such as packet loss, interface speed, errors, inbound traffic and more are rendered as interactive trends, turning raw counters into insight for troubleshooting and capacity planning.
Discovered nodes and links plotted on a live geographic map, giving operations an end-to-end, spatial picture of the network and its service paths at a glance.
The Impact
The same operations team, working a fundamentally better way.
| Dimension | Before: siloed & manual | After: unified & automated |
|---|---|---|
| Visibility | A separate EMS per vendor, with no unified view | One pane of glass across ~27,000 devices and 3 layers |
| Alarm handling | Alarm floods with no correlation or deduplication | Correlated incidents cut alarm noise and speed root-cause analysis |
| Ticketing | Manual root-cause analysis and manual ticketing | Automatic incident-to-ticket into HPSM |
| Service turn-up | CLI-driven, device-by-device provisioning | Catalog-driven automation for 5 service types, with rollback |
| Inventory | Static, fragmented inventory records | Live graph inventory plus geographic topology |
| Scalability | Tooling that couldn't scale with the network | Cloud-native platform built to grow with the network |
Fault resolution
Correlation, enrichment and auto-ticketing compress detection-to-resolution and shrink outage windows.
NOC operations
Multi-algorithm correlation and deduplication surface the real problems and silence the noise.
Service turn-up
Catalog-driven provisioning replaces manual CLI work with consistent, repeatable, low-risk workflows.
Network visibility
Fault, performance, inventory and topology in one place, across every vendor and every layer.
Capacity & planning
KPI thresholds, trends and analytics move operations from reacting to anticipating.
Scalable foundation
Cloud-native, highly available and open-standard, ready for the DWDM and legacy phases and beyond.
A single, open OSS that turns a fragmented, multi-vendor network into one observable, automatable system, giving operations the visibility and control to run at national scale.
One cloud-native OSS for fault, performance, inventory, topology and catalog-driven fulfilment, ingesting from any vendor, integrating with the ERP and ticketing systems you already run.
A network of hundreds of thousands of sites generating ~1.25 crore alarms a day, handled by static correlation rules and manual RCA. Six coordinated AI capabilities turned that flood into unique actionable incidents, without a single byte leaving the operator's estate.
Thousands of change requests a month across hundreds of change categories, assessed by checklist and delivered by hand. Seven coordinated flows turn each CR into a risk-scored, simulation-backed, auditable operational outcome.
O-RAN gives operators vendor choice, and with it more moving parts to coordinate on every site turn-up. One automation layer now carries planning data, field capture, provisioning, configuration and verification end to end.
Customer identity withheld and referred to throughout as "one of India's largest neutral telecom infrastructure providers." Figures and screens are sourced from the NetSingularity OSS solution design; customer-identifying data redacted.
