Intuceo built the Smart Correlation Tool (SCT) on Palantir Foundry to bring alarms, trouble tickets, network telemetry and customer-reported issues into one correlation layer. Using a network ontology that acts as a digital twin of the network, SCT traces multiple downstream issues to a single validated root cause and manages them as one incident.
- Industry Telecommunications and network operations
- Client type Network operator with multi-team operations
- Engagement Incident correlation, root cause identification and managed incident workflow
- Core stack Palantir Foundry, network ontology (digital twin), near real-time data ingestion, custom graphical user interface (GUI), application programming interface (API) layer
4
Data streams correlated: alarms, trouble tickets, network telemetry and customer-reported issues
1
Managed incident per validated root cause, replacing multiple independent tickets
12
Core capabilities, from real-time correlation to noise suppression
2
Access modes: an operational GUI and an API layer for operations and business support system (OSS and BSS) integration
Client Overview
A network operator managing alarms and tickets across organizational boundaries
The client runs telecom network operations in which alarms, trouble tickets and customer complaints arrive through separate systems and are handled by separate teams. Property, network operations and field teams each see part of the picture, but none of them sees how a single upstream failure spreads across circuits, services and customers.
The Challenge
Disconnected events, and no single view of the cause
Without a correlation layer, the client’s teams were working from reactive troubleshooting rather than proactive incident management. Five problems stood out.
- Uncorrelated volume: High volumes of alarms and trouble tickets arrived with no link between related events.
- Duplicated effort: Multiple teams troubleshot the same underlying issue independently.
- No upstream view: Teams could not tell upstream failures from downstream symptoms.
- Unnecessary dispatches: Field crews were sent to issues an existing failure already explained.
- Customer disconnect: Customer complaints were hard to link to actual network faults.
The Solution
An ontology-driven correlation layer that finds one root cause for many symptoms
Intuceo built SCT on Palantir Foundry as a correlation and incident management layer that sits across the client’s existing systems. At its core is a network ontology, a digital twin that models how fibers, circuits, services, equipment and customers connect. Each incoming record is evaluated against that model to decide whether it is isolated or part of a wider failure.
- Topology traversal: When a defect is detected, SCT maps it to the ontology and walks the hierarchy from fiber to circuits, services and customers to find every downstream impact.
- Root cause validation: Topology traversal and rule-based validation confirm a shared upstream dependency before escalation, which keeps false positives out of the queue.
- Single managed incident: All related defects, alarms and tickets are linked to one incident with unified ownership.
- Automated routing: The managed incident goes straight to the right team, such as property, network operations or field crews.
- Noise suppression: Duplicate downstream alarms and tickets are suppressed once the upstream failure is known, reducing alert fatigue.
- Historical intelligence: Every relationship between defects, incidents and network entities is stored for trend analysis and recurring-issue detection.
Technology used
| Layer | Technology | Role in this engagement |
|---|---|---|
| Data foundation | Palantir Foundry | Hosts the correlation logic, ontology and managed incident objects |
| Network model | Network ontology (digital twin) | Models fiber to circuit to service to customer relationships |
| Ingestion | Near real-time data streams | Brings in alarms, trouble tickets, telemetry and customer issues |
| Correlation and validation | Topology traversal with rule-based validation | Identifies and confirms the most probable root cause |
| User access | Custom GUI | Lets operators explore incidents, relationships and topology |
| Integration | API layer | Connects to OSS and BSS and client-facing applications |
How It Works
Ingest and map
Alarms, tickets and customer issues are ingested and mapped to the network ontology.
Trace the impact
SCT computes the topology-down impact. A fiber cut, for example, maps automatically to every circuit riding on that fiber.
Correlate and validate
Related defects are grouped and the shared root cause is validated.
Manage and resolve
One managed incident is created, routed and tracked through the GUI and APIs until closure.
Results and outcomes
Faster resolution, fewer dispatches and less alert noise
- Reduced mean time to resolution (MTTR): [X]% reduction against the pre-SCT baseline
- Lower operational cost through dispatch optimization: [X]% fewer unnecessary field dispatches
- Reduced alert noise and duplication: [X]% reduction in duplicate alarms and tickets
- Improved cross-team collaboration: Teams work one managed incident instead of separate tickets
- Better customer experience: Accurate impact mapping supports clearer customer communication

