Skip to main content
The Spine Fabric continuously ingests signals from your support stack, normalizes them into a shared model, and projects domain-specific views on demand. For Support, this means a ticket, its customer history, an active incident, and linked engineering work all appear together without manual linking.

​
Sources

  • Zendesk / Intercom / Freshdesk: Tickets, customer records, macros, and SLA policies.
  • Slack: Thread references, incident channels, and on-call alerts.
  • Linear / Jira: Engineering tasks created from escalations and their current status.
  • Sentry: Error events and stack traces linked to reproductions.
  • PagerDuty: On-call schedules, incident pages, and escalation policies.

​
Scopes

Support data is scoped by account, assignee, queue, priority, and time window. For example, an on-call responder sees incidents plus related tickets; a support manager sees SLA dashboards across the full team.

​
Event Sequence

  1. A ticket or incident is created in a source system.
  2. Spine Fabric maps it to the unified entity model (Ticket, Incident, Customer).
  3. Twin and core agents identify related entities and suggest context.
  4. The projection is assembled and rendered in the relevant Support view.
  5. User actions update state in Spine and write back to the source system.

​
Projections

Projections are query-time compositions. A Customer 360 projection assembles tickets, contacts, and events from three tools by account ID. An Active Incidents projection joins PagerDuty incidents with related tickets by time and tags.