27,000 network devices, one governed operations centre
A carrier-scale Tier-1 operator in India brought more than 27,000 routers, switches and devices under one governed operations centre with x101, cutting mean time to resolve by a measured 43% and automating 85% of routine manual operations. Fault and performance data that had arrived over three separate collection paths now lands in a single near-real-time layer, where engineers see investigated insight, ask for live values in plain language, and act through governed automation.
Telecom · India
Last updated: 23 September 2026
The challenge
The operator's estate, more than twenty-seven thousand devices, generated a continuous flood of fault and performance data over three different collection mechanisms. Nothing monitored those devices centrally in real time, so what was happening inside the network was fragmented across collection silos and the operations centre saw it late, partially, or not at all.
No central real-time view
Twenty-seven thousand devices, and no single place to see their fault and performance state as it happened.
Three ingestion worlds
Traps and metrics, event streams and polled collection each lived in its own pipeline, with no correlation across them.
Volume beyond human triage
Carrier-scale telemetry arrived faster than any team could read it, and meaningful signals drowned in the stream.
Investigation as archaeology
Root-causing a fault meant hopping across element managers and counters, device by device.
Dashboards without answers
Views, where they existed, showed symptoms; a live value still meant logging into the device.
Network, infrastructure and services apart
Device faults, server health and service state were monitored in separate tools with separate truths.
What x101 does
x101 was deployed as the central intelligence layer for the operations centre: ingesting all three collection paths at carrier scale, processing fault and performance streams in near real time, and putting investigation on the platform rather than the engineer. Probable causes are suggested with evidence, approved procedures run automatically, incidents are raised with context, and engineers ask for live values in plain language instead of hunting through consoles.
Three paths, one pipeline
Traps and metrics, event streams and polled collection from 27,000+ devices land in one pipeline, normalised and correlated on arrival.
Processed as it arrives
Events and performance indicators are handled as they land, so degradation surfaces within moments rather than at the next reporting cycle.
Investigated, not just displayed
Each fault is investigated against topology, history and correlated signals, and the probable cause is presented with its evidence. The engineer decides.
Live values, asked for in words
Engineers ask for current values and states in plain language and get answers drawn from the live estate, alongside the central dashboards.
Known patterns, handled
Approved procedures run automatically on recognised patterns, incidents carry their context, and every action is reversible and audited.
Network, infrastructure, services as one
The same layer watches the infrastructure and the services riding on the network, so cause and impact are visible in one place.
Governed autonomy, deliberately bounded. x101 suggests root causes rather than acting on its own judgment, automation is limited to predefined and approved procedures, and every incident, ticket and action carries a full audit trail and a rollback path. Autonomy is earned scope by scope, never assumed.
The impact
The operations centre moved from fragmented, delayed device monitoring to one near-real-time picture of twenty-seven thousand devices, with investigation, live values and action available in the same place the fault appears.
- 43%
Lower MTTR
Measured in production within six months of go-live.
- 85%
Routine ops automated
Running as governed procedures rather than manual work.
- 27,000+
Devices in one view
Three ingestion paths correlated into a single pane.
- ~60%
Incidents auto-created
Raised automatically with probable cause and context attached.
| Dimension | Before | After · with x101 |
|---|---|---|
| Device visibility | No central real-time view across the device estate | 27,000+ devices in one near-real-time fault and performance view |
| Data ingestion | Three collection paths in separate, uncorrelated silos | One pipeline, normalised and correlated on arrival |
| Investigation | Manual, device-by-device root-causing across consoles | Probable cause suggested with evidence attached, engineer in command |
| Access to truth | Log into the device to read an actual value | Live values and states answered in plain language |
| Response | Ad-hoc manual procedures, untracked | 85% of routine operations automated as governed procedures, 43% MTTR reductionBoth measured in production within six months of go-live |
| Coverage | Network, infrastructure and services in separate tools | One governed layer across devices, infrastructure and services |
At twenty-seven thousand devices the question is never whether the data exists, it always does. The question is whether anyone can see it in time, trust what it means, and act on it safely.
Solution summary · x101 carrier-scale network observability deployment
The parts of x101 this uses
Nothing here was built for one customer. Each capability below is standard platform behaviour, applied to this problem.
Unify
Three collection paths, plus infrastructure and service state, read as one correlated picture.
Explore UnifyAgentic Orchestration
Topology, history and signals investigated in parallel, converging on one probable cause.
Explore Agentic OrchestrationInstant Analytics
Live values and performance questions answered from the estate as it is right now.
Explore Instant AnalyticsFrom Answer to Action
Approved procedures execute and incidents are raised with context, inside the same flow.
Explore From Answer to ActionAI Explainability
Every suggested root cause arrives with the evidence behind it.
Explore AI ExplainabilityGoverned by Design
Bounded autonomy, audited actions and a rollback path on everything.
Explore Governed by Design
About this case study. The customer's identity is withheld at their request and is referred to throughout as a Tier-1 telecom operator in India. The MTTR reduction and routine-operations automation figures are measured production results from this deployment; the remaining figures are indicative, and actual results vary with estate scope, telemetry profile and rollout phase.
Bring your device estate into one governed operations centre.
Carrier-scale ingestion, near-real-time processing, investigation with evidence, and governed automation, deployed inside your environment.