A knowledge base is a static repository your agents search by hand. Knowledge Intelligence is an active layer that knows which article applies to which issue, notices where you have no answer at all, and drafts the missing one from your own files.
Knowledge base authoring is slow, unrewarding work that competes with closing tickets. Given the choice, everyone closes the ticket. So the same five questions escalate every week, and the article that would have stopped them never gets written.
A wiki can only tell you what you know — a gap looks exactly like an empty search result. Knowledge Intelligence detects where no solution exists at all, and Top Call Drivers rank those gaps by the volume still arriving.
Title, template, sources, draft.
Fixes the structure and controls what gets generated in each section, down to the columns in a generated table.
Metadata that organises the article for search and retrieval, so generated knowledge lands somewhere findable.
Upload source files or select data sources already connected. The agent learns from them and drafts from the relevant content.
Adjust for accuracy and tone, enrich with GenOnFly, restyle with the styling agent, then publish and export.
What separates this from asking a chatbot to write your documentation.
Generation runs against files you supply or have already connected, so the draft comes from your own manuals, tickets and documentation rather than the open internet.
“The knowledge agent learns from these files and will automatically process the relevant content to generate a draft.”
A template is not only a shape. Each section carries its own generation instructions, defined once and applied to every article built from it.
“Not only is the structure of the article defined in template, we also have the flexibility to control the content that will be generated in each section.”
A starting point a person refines before it goes anywhere. Nothing is auto-published, which for a knowledge manager is the whole conversation.
Each solves part of it and leaves the expensive part behind.
A capable repository with no gap detection, no generation and no usage feedback. It stores what you give it.
The right methodology, but the authoring is manual, so articles lag reality by months and the backlog only grows.
Not grounded in your corpus, with no template control, no review routing and no path to publish. Prose, not a workflow.
Terms used in this recording and its companion KCS walkthrough.
The longer companion tour, going deeper on personas, the styling agent and GenOnFly.
The product page: gap detection, coverage, and the leadership view of what knowledge is worth.
The other half of the loop, where a resolution found in search becomes a published article.
Knowledge Intelligence
Most teams cannot say what fraction of their articles were updated in the last year. We will look at your corpus and show you where the gaps are.
Talk to us