Corporate RAG pipelines

Governed knowledge bases and multi-tenant RAG with hard isolation for internal copilots.

Platform teams use Synamcps as the governance layer for corporate RAG: ingest PDFs and wiki links into per-team storages, let extraction → summary → embeddings run automatically, and expose retrieval via POST /api/knowledge/search or knowledge_* MCP tools.

Document visibility (personal / group / public) sits on top of storage ACLs. HR group docs stay out of an engineer's results even when both share a storage — necessary for multi-team copilots and multi-tenant isolation with tokens scoped to storageIds and maxMode: read.

Wire Confluence, Notion, or Jira exporters to POST /api/knowledge with stable source and sourceUrl labels. Re-ingest updates idempotently; filter retrieval by source to scope queries. The SSRF guard blocks attacker-supplied links from pivoting to cloud metadata.

Every call flows through tokenized auth with Prometheus metrics and Redis TimeSeries usage events, so security can answer which agent read which storage without a separate observability stack.