AI Employee Infrastructure

The Brain

The shared library every AI employee draws from. Sales, ops, marketing, support, and research agents read the same curated context — with hierarchy, permissions, and poisoning prevention built in.

Governed Context Layer

Centralized definitions and metadata — agents pull from a single verified source of truth.

Context State Graph

Incremental memory with who changed what, when, and why. Latest nodes surface first.

Hierarchy & Permissions

Access enforced per person and role. Top-level actors define order of operation for the org.

Context Drift Detection

Monitors glossary age, schema staleness, and ownership freshness to catch metadata decay.

Strict Data Lineage

Full traceability from upstream source (Slack, Notion, Gmail, MCP) to consumed context.

Automated Invalidation

Cuts off or flags stale context the moment an upstream definition changes.

API Endpoints

POST/api/brain/searchGoverned search for agents (semantic + incremental)

POST/api/brain/ingestIngest context nodes, definitions, invalidations

POST/api/integrations/slackIngest Slack conversation outcomes

GET/api/brain/nodesList or fetch context graph nodes

GET/POST/api/brain/driftView or run context drift scans

GET/api/brain/lineageTrace data lineage for a node

GET/api/brain/violationsPermission violations and override attempts

GET/api/brain/healthService health check