Look at the row under Website pages. A failed pill with · error next to it means the job finished with nothing to store. Hover · error to read the reason. The usual reason reads: Crawl returned no readable content — the site may be blocking automated crawlers. Add content manually or retry. Click the bin icon on the row, click Confirm delete, then add the address again. If it fails a second time your site is turning the crawler away, so put the wording in a custom reply instead. The row reads ready and Last crawled shows a time just now.

Teach it your business

Add your website pages to your bot

When you finish, your bot answers visitors using the pages on your own website.

Before you start

You have a bot in your Adoomi dashboard How to check
Your website is live at a public address anyone can open
You know which pages hold the answers — prices, opening hours, services

About 5 minutes. You’ll do four things:

pick a whole site or one page
paste the address
wait for the row to read ready
ask your bot a question

Your bot answers from what you give it, and your website is the quickest thing to give it. The answers are already written there. You paste an address, Adoomi reads the pages, and the bot starts using them.

Two ways to add pages. Pick before you start.
Option	Best for
Crawl site	Your whole site, or a whole section of it. Adoomi starts at the address you give and follows the links it finds.
Single URL	One page you know you need, such as a price list. Adoomi reads that page and nothing else.

Add the pages

In the bot sidebar, under Bot setup, click Train your bot.
The Web pages tab opens first.
Under the address box, next to Mode, choose Crawl site or Single URL.
Type your address into the box that reads docs.your-business.com or a single page URL.
Adoomi puts https:// on the front for you, so you can leave it off.
Click Crawl & index.
The button reads Fetching… while the job starts.
Wait for the new row under Website pages and watch its Status.
A typical crawl finishes in 1–3 minutes, and the row updates itself while you watch.

How far a site crawl reaches

Crawl site starts at the address you gave and follows links outwards. It stops at 100 pages, and at three links deep from that starting address. Neither number can be changed from the dashboard. Adoomi keeps the main content of each page and leaves the navigation and footer behind.

A site crawl also skips address patterns that are usually noise on a website:

tag pages, such as example.com/tag/boilers
category pages, such as example.com/category/news
search result pages
printer-friendly pages
feed pages
AMP pages

If one of those is a page you actually need, add it on its own with Single URL. That route fetches exactly the address you paste and skips nothing.

Reading a row

What the Status pill tells you
Status	What it means	What to do
pending	Adoomi is still reading the site.	Wait. A typical crawl finishes in 1–3 minutes.
ready	The pages are stored and the bot can answer from them.	Ask your bot something only those pages answer.
failed	The job finished with nothing worth storing.	Hover the · error beside it to read why.

A second badge appears beside the status only when something is wrong with the address itself. Adoomi checks your web-page addresses once a week, on Monday mornings at 06:00 UTC, and hovering the badge tells you when it last looked.

What each health badge means
Badge	What Adoomi found	What to do
404	The page was not there at the last check.	Delete the row and add the current address.
Redirect	The address now sends visitors somewhere else.	Add the new address, then delete the old row.
Auth required	The page asks for a login.	Adoomi cannot read it. Put the wording in a custom reply instead.
Server error	Your site returned an error to the check.	Open the page yourself. This one is often temporary.
Timeout	The page took more than ten seconds to answer.	Open the page yourself, then retrain that source.
Unreachable	Adoomi could not connect to the address at all.	Check the address is public and spelt correctly.

If it didn’t work

The row says failed

Check: Look at the row under Website pages. A failed pill with · error next to it means the job finished with nothing to store. Hover · error to read the reason.
Fix: The usual reason reads: Crawl returned no readable content — the site may be blocking automated crawlers. Add content manually or retry. Click the bin icon on the row, click Confirm delete, then add the address again. If it fails a second time your site is turning the crawler away, so put the wording in a custom reply instead.
Resolved when: The row reads ready and Last crawled shows a time just now.

The address was refused

Check: The box shows That URL isn't allowed (must be a public website). Read back the address you typed. Adoomi refuses anything that is not a public website: a local address, a private network address, or a single word with no dot in it.
Fix: Type the public address of the page instead, such as example.com/prices. Check it first by opening it in a private browser window. If your site is not published yet, publish it and add the page afterwards.
Resolved when: A row appears under Website pages and its Status starts at pending.

A page I need is missing after the crawl

Check: Type part of the address into Search pages and see whether it has a row. A site crawl stops at 100 pages and three links deep, and skips tag, category, search, print, feed and AMP addresses.
Fix: Add that page on its own. Next to Mode, choose Single URL, paste the full address of the page, and click Crawl & index.
Resolved when: The page has a row of its own under Website pages, reading ready.

The row has been pending for a long time

Check: A typical crawl finishes in 1–3 minutes. Leave the tab open, because the row refreshes itself.
Fix: If it is still pending after ten minutes, reload the page. A crawl whose worker has stopped is marked failed within about ten minutes; one that is still running is left for up to an hour. Once the row reads failed, delete it and add the address again.
Resolved when: The row reads ready, or failed with a reason you can act on.

The bot answers with the old wording from that page

Check: Read Last crawled on the row. Adoomi stored what the page said at that moment, not what it says now.
Fix: Open the Retrain tab and click Update bot with latest. Wait for the row's Status to go back to ready, then ask again.
Resolved when: Last crawled shows a time just now and the answer matches the page.

None of these fixed it

Check: You have checked the status, the badge and the address, and the bot still does not answer from the page.
Fix: Email hello@adoomi.ai with three things: the page address, your bot's name, and what the row's Status reads.
Resolved when: A person at Adoomi replies with the cause and what to change.

What this can’t do

A site crawl stops at 100 pages and three links deep from the address you gave. Neither number can be changed from the dashboard — add the rest with Single URL.
Pages behind a login are not supported. Adoomi's crawler carries no password and no cookies, so it sees only what a stranger sees.
Local and private addresses are not supported. localhost, a private network address and a single-word hostname are all refused.
A site crawl skips tag, category, search, print, feed and AMP addresses. Add one of those with Single URL instead.
Adding an address that is already in the list replaces that row rather than adding a second one, and the Mode you picked this time wins.
Web pages you add here read 0 B against Total size in the right-hand rail. That allowance counts documents, FAQs and custom replies, not pages crawled from this tab.
Deleting a row removes those pages from what the bot knows straight away. Adding the address back starts a fresh read.

Healthcare

Channels

Help Center