Knowledge agents
Agents that answer from the organization's own regulations, procedures and records — with the source attached.
What does our policy say about this case, and which clause says it?
The answer, the clause it came from, and the document and version it was read from.
What a knowledge agent is designed to do
Every organization has a written answer to most of the questions it is asked, and no reliable way to reach it. The regulation is current, the circular that amended it is not filed with it, and the person who remembers both is on leave. A knowledge agent is the one that will read the written record rather than paraphrase it.
The pack is designed to:
- answer with the clause, the document and the version it was read from;
- say when two documents disagree, instead of choosing one silently;
- say when the written record does not answer the question at all.
Agents answer from your own regulations, documents and records, with the source attached.
The primitives it composes
An agent is a model, a set of approved skills, tools and a policy, applied to a scope.
Agents reach your systems through approved connectors — APIs, databases, applications — each with its own permissions.
Changes wait for human approval before they run.
What it runs on
Skills come from a signed vendor catalogue or your own uploads, and can be enabled or disabled per workspace.
Sign in with your own identity provider (OIDC with PKCE); no shipped default administrator password.
Bring your own model — OpenAI-compatible endpoints, Ollama, Anthropic, or Cloudflare AI Gateway / Workers AI — with keys envelope-encrypted at rest.
The whole stack self-hosts with one command; an air-gap bundle ships every image.
Bring one question of your own.
Request a demo