Skip to content

Core Knowledge

Structured knowledge that makes AI answers accurate, complete, and source-aware.

Upload documents, spreadsheets, directories, policies, websites, and internal resources so Core Guides can answer from information your organization controls.

Documents
Tables
Assignments
Retrieval testing
Gap review

Knowledge handling

Different sources need different treatment.

Auto handling

Let Core detect whether a source is a standard document, structured table, exact reference, or keep-together source.

Knowledge assignments

Assign sources to the Guides that should use them so each Guide answers from the right library.

Structured lists and tables

Directories, rosters, pricing, contacts, programs, and spreadsheets stay more useful in answers.

Exact-reference content

Important facts like emails, phone numbers, requirements, and deadlines can be preserved carefully.

Source priority

Important knowledge can be weighted so Guides prefer authoritative sources.

Completeness-aware answers

Guides can avoid partial answers when a complete set of details is required.

Knowledge Gap Queue

Turn unanswered questions into better knowledge.

When a Guide cannot answer from assigned knowledge, Core can capture the question in a Knowledge Gap Queue. Administrators can review the gap, use an AI-proposed answer as a starting point, edit or replace the answer, and create a new knowledge source.

Capture gaps automatically

Core can record questions that could not be answered confidently from assigned knowledge.

Review proposed answers

AI can draft a proposed answer so administrators are not starting from a blank page.

Approve, edit, or replace

Teams can approve the draft, edit it, replace it, ignore the gap, or mark it as a duplicate.

Create clean sources

One gap can become one source, or multiple same-Guide gaps can become one combined FAQ source.

Assign back to the Guide

New sources can be assigned to the originating Guide by default.

Improve continuously

Guides become more accurate as real user questions reveal what knowledge is missing.