1
Intake
Spine polls or receives webhooks from HubSpot, Marketo, Salesforce, Google Analytics, LinkedIn Ads, Segment, and Notion.
2
Load
Raw records are ingested into Spine staging with source metadata and ingestion timestamp.
3
Normalize
Records are mapped to the Spine entity schema: Campaign, Audience, Content, Lead, Attribution, Channel, Program, and Event.
4
Schema Mapping
Cross-system matching links UTM parameters to campaigns, emails to leads, and page paths to content assets.
5
Context Assembly
Spine builds campaign graphs by joining performance metrics, lead flow, content assets, and revenue data.
6
Projection
The assembled graph is filtered and shaped into domain views: Campaign Performance, Attribution Funnel, Content Pipeline.
7
OODA
Twin observes projected state, orients to campaign health models, and decides whether to surface an alert.
8
Governance
Every projected view respects source-system permissions and domain guardrails before rendering.
9
Execution
When you invoke a capability, Spine routes the call to the correct connector with scoped authorization.
10
State Update
Writeback responses update the local projection, trigger events, and refresh dependent views.