1
Intake
A source system emits a raw event: a ServiceNow ticket update, a Slack message, or a scheduled HR feed export.
2
Load
Spine Fabric receives the event through a connector and records it in the raw event log with source timestamp and connector identity.
3
Normalize
The event is parsed into a common schema. Field mappings translate ServiceNow status values, Jira issue types, and Slack channel names into shared enumerations.
4
Schema Mapping
The normalized event is matched to entities: a new ticket maps to a
Ticket and optionally to a Request, Asset, or Vendor.5
Context Assembly
Spine Fabric traverses relationships to pull the submitter profile, related tickets, asset history, vendor record, policy rules, and open approvals into an entity graph.
6
Projection
The enriched graph is written into the relevant domain views: Request Queue, Asset Inventory, Vendor Portfolio, or Approval Dashboard.
7
OODA
Twin observes the projected change, orients it against known patterns, and decides whether to alert a practitioner, propose an action, or delegate to an agent.
8
Governance
Every projection update is checked against authorization rules. Unauthorized or anomalous patterns trigger a governance alert.
9
Execution
When a capability is invoked, the execution engine resolves the target system, translates the action back to source-specific parameters, and dispatches the call.
10
State Update
Source systems return new state, which re-enters the pipeline at Intake so views stay eventually consistent.
Each stage produces a trace record. You can inspect the full path of any entity from raw event to projected view.