Where your knowledge lives now
You do not attach content to a chatbot any more. Everything the AI can read belongs to the brand, on one screen: Brand → Knowledge. Whatever is listed there is searched by the brand's chat AI, by its phone agents, by the reply suggestions your team sees in the inbox, and by AI social posting — in a single search, with nothing to connect up.
One screen, two halves
The top half is your structured content: knowledge bases and documentation sets. These are things your customers can also read, if you choose to put them on your public web address.
The bottom half is Other sources — everything else the AI reads, and nothing your customers ever see. The page says as much on screen: "Everything else your AI reads. None of this appears on your public site — it is knowledge only, used to answer in chat, on the phone and in social posts." The Add button there offers four things:
- Paste text — anything you find yourself repeating.
- Question & answer — one question, one answer.
- Upload files — PDF, DOCX or TXT.
- Website page or sitemap — give us the address and we read it for you.
A second tab beside Sources holds API tools, covered further down.
Publishing indexes it straight away
When you publish a help article, it becomes searchable by the AI immediately. There is nothing to attach, no chatbot needed, and no widget needed. Edit an article and only that article is re-read; unpublish it and it is removed from what the AI can find. If you have used Yaplet for a while, this is the change most worth knowing: content that could not be traced from a widget to a chatbot used to be silently never indexed at all.
A related trap is also gone. Swapping which knowledge base a brand shows, moving one to another brand, or taking one down no longer leaves the old content answering. It used to stay indexed for ever, so the AI kept quoting articles that had been taken away.
Status and Indexed — the two columns worth learning
Every row carries a Status, and it is the fastest way to find out why a question came back empty:
- Ready — stored and searchable.
- Not indexed — nothing was stored from it, so the AI cannot answer from it.
- Reading… — still being processed.
- Failed — we could not read it. Hover for the reason.
- Warning — read, but not cleanly. Hover for the reason.
Beside it, Indexed shows how many searchable pieces the AI stored from that source. A dash means nothing was stored. One article usually becomes several pieces, and only published ones count. You can filter the table by status, and if anything failed its last read you get a warning as soon as the page loads.
Which knowledge base is on your public site
A brand can hold up to five knowledge bases and five documentation sets on any paid plan (one of each on the free plan), and the AI reads all of them. But only one knowledge base and one documentation set can sit on the brand's public web address at a time. The Public site column shows which, reading Live or Hidden, and the row menu offers Show at (your address), Visit the live page and Take down from the public site.
Taking one down removes the public page and nothing else — the articles are untouched and the AI keeps answering from them. The first knowledge base you create in a brand goes live automatically; later ones start out Hidden. If the brand has no web address at all, a notice at the top of the page says so and offers to set one.
One word of vocabulary, because it saves confusion: in Yaplet, publish means one thing only — the per-article switch that decides whether that single article exists. For a whole knowledge base the words are Live, Hidden, Show at… and take down.
What language a source is written in
Each knowledge base and documentation set has its own Language, set inline on its row. The screen explains it: "The language this is written in. It decides how your AI searches the text — it does not translate anything." The brand's own Language, in Brand settings → General, does two jobs: it is the fallback for things that have no language of their own — files, pasted text and question-and-answer entries — and it always takes the first of the three keyword-search slots, whether or not anything depends on it.
Copy or move something to another brand
If you run more than one brand, the row menu has two more options. Copy to another brand… duplicates the whole thing — every category, article or page, with its nesting, language and ordering — and everything arrives unpublished, so nothing goes onto the other brand's public site, nothing is indexed and nothing is billed until you publish it there. Move to another brand… hands the original over with everything it has already learnt, costs nothing, re-indexes nothing, and arrives Hidden. If you are moving the one currently on this brand's address, the dialog warns you in red that the address will have nothing behind it afterwards.
Deleting a source
Delete is a single action with a counted confirmation — it tells you exactly what goes ("84 articles and their categories") and that everything indexed from it goes too. If you are deleting the source currently on your public address, it says that as well. There is no undo.
Smart import — let us read your website
Beside the Add button is Smart import. Give it your website address and we read the site's navigation or sitemap, pick the couple of dozen pages that actually help a support AI, and import them in the background — a few minutes, usually. Running it again updates pages we have read before instead of creating duplicates.
Re-reading a website page
Use Read again now in the row menu to re-read a page immediately. A page can only be read again once every five minutes. Yaplet also re-reads website pages and sitemaps by itself roughly once a week — but only for organisations on a paid plan, because re-reading costs real processing. On a free account, use Read again now when the page changes.
API tools for live data
Some answers cannot come from stored text — an order status, current stock, a booking. For those, an API tool lets the AI call one of your own endpoints while it is writing the answer. The column that used to be called Description is now When it is used, and the wording matters: the AI reads that text to decide when to call the tool, so describe the situation, not the endpoint. Each tool reads Active or Paused.
The allowance is per brand: one API tool on the free plan, twenty on paid plans. Names may use capital letters and hyphens, up to 64 characters, and a name that only differs in capitalisation is rejected as a duplicate.
The 250 indexed-item allowance
On the free plan a line between the two halves of the page reads something like "212 of 250 indexed items used". That allowance is shared across your whole account — every brand, knowledge base, documentation set, file, website page and question-and-answer entry together. Remember that one article is usually several indexed items, so 250 is not 250 articles. When it is full, publishing new articles and adding new sources is paused until you remove something or upgrade. The same 250-item allowance also applies if you bought individual modules rather than a plan — buying a module buys features, not content volume.
Who can open this screen
The page opens for anyone holding any one of three permissions: Vex, Knowledge base or Documentation. Creating or changing a knowledge base needs Knowledge base; a documentation set needs Documentation; and the whole Other sources half is only shown to people who hold Vex.
Product feeds are on their own screen
Product catalogues are deliberately not listed here — they live only at Brand → Products, so the same thing cannot be deleted from two places. A brand with no AI agent cannot add one yet and shows "Add an AI agent first" instead. Yaplet re-reads a product feed about once a day for organisations on a paid plan.
Next step
Check that what you added actually reached the AI: Test your chatbots before going live. Once real conversations have happened, see what is still missing: Find knowledge gaps Vex couldn't answer.