Anetworkestateonemodelcanreasonabout
Telecom operators run one of the largest and most instrumented asset estates in any industry, across systems that were never designed to describe each other.
Theproblemiscorrelation,notdatavolume
A network operations centre is rarely short of signals. It is short of correlation: an alarm storm, a customer complaint pattern, a scheduled works notice and a capacity trend are four views of one situation, held in four systems with four different identifiers for the same site.
That is an ontology problem before it is an AI problem. Once a site, its equipment, its dependencies, its planned works, its incidents and its served customers are one object, both correlation and impact analysis become traversal rather than investigation. Agents then become useful because there is something for them to reason over.
Thenetworkontology
A network element, its site, the outage that affects it, the work order raised, the engineer dispatched and the agent authorised to prepare the case — four systems' worth of identifiers, one object model.
- Field engineer—ASSIGNED_TO→Work order
- Dispatch agent—AUTHORIZED_FOR→Work order
- Work order—RELATES_TO→Network element
- Network element—LOCATED_AT→Site
- Site visit—INSPECTS→Network element
- Site visit—PRODUCES→Evidence
- Outage—AFFECTS→Network element
Whereitapplies
Network operations
Correlating alarms, changes and customer impact into one situation with a proposed next step.
NOC support
Assembling the case for an operator: what changed, what depends on it, what has worked before.
Capacity planning
Bringing utilisation trends, planned works and commercial commitments into one view.
Field service
Dispatch and on-site support with the assembled site history and applicable procedure.
Asset management
Equipment lifecycle, condition and dependency across a distributed estate.
Incident handling
Impact analysis, communication drafting and consistent post-incident records.
Cyber operations
Investigation support over asset, identity and network context with evidence retained.
Infrastructure planning
Build and upgrade prioritisation against topology, demand and constraint.
Knowledge access
Procedures, configurations and prior resolutions, findable with citations.
Network configuration changes go through existing change management with a qualified engineer. The platform prepares and evidences the change; it does not push configuration to production network elements unattended.
Questions
- Does this replace our OSS or BSS?
- No. Those remain the systems of record. The ontology defines what their objects mean in relation to each other, which is the part that is currently done by people holding several screens open.
- Can agents act on the network?
- Not unattended. Proposals go through existing change management. In practice the value is in the time between an event and a well-prepared decision.
- How does this handle scale?
- The ontology is modelled by type rather than by instance, so a large estate is a data-volume question for the runtime rather than a modelling question. Start with one operational decision and one region.
Bring us one operational problem.
You do not need a finished brief. Bring the problem — we will work out the next step together.
Or book a call with the team