Connect organizational knowledge

What's organizational knowledge?

Your company's own information — from Slack, email, documents, tickets — gathered into a searchable index. With it, an assistant answers from real company facts, with citations, instead of guessing.

How it works

• OpenCrane runs collectors that continuously pull in knowledge from your systems and organize it by scope (personal, project, department, org).
• During a conversation, an assistant looks things up directly and answers with citations. OpenCrane never reads the conversation — it only decides which knowledge an assistant is allowed to see.
• What an assistant can reach is set by access: a department's documents only reach that department.

Keep every assistant consistent

So every assistant behaves the same way when it looks things up — same rules for which sources to use, when to cite, and how fresh information must be — OpenCrane applies a shared set of rules across the fleet. You roll changes out gradually (to a few assistants first, then everyone) and can undo in one step:

oc awareness rollout show
oc awareness rollout set <version> --waves engineering,sales
oc awareness rollout promote
oc awareness rollout rollback

Keep contexts from bleeding together

To stop one project's context from leaking into an unrelated chat, you can pin a conversation to a scope:

oc sessions scope set <session> --principal alice --scope project:acme
oc sessions scope show <session>
oc sessions scope clear <session>

Going deeper

How collection, datasets, and freshness work under the hood is in the Retrieval & memory deep dive. Health dashboards are in Awareness SLOs.