A site chatbot that can't make things up.
Paste your URL. In about a minute there's a working assistant on your pages, and a visitor who asks in plain words lands on the page that answers them instead of hunting your menu. No model writes the reply, so there is nothing for it to invent. It routes to something you published, or it declines. You do nothing.
No sign-up, no card. We read your public pages and build a preview, hosted here.
Here's your bot - on your own content.
Put it on your site →
We'll build yours and send you the link.
Drop an email and we'll send your bot's link.
Most site chatbots can say anything. This one can only point to your own pages.
A website answer engine that cannot make things up
Most chatbots write their answers, which is exactly why they can make things up. This one never writes anything. It takes the visitor’s question, finds the page on your site that answers it, and sends them there. If the question is vague, it offers a short list to choose from. If nothing on your site answers it, it says so plainly, instead of inventing something to fill the gap. And because no model runs when it replies, you get the same answer every time, and you can see exactly which page it picked and why. It can still pick the wrong page, but it will never make up a fact that isn’t on your site.
Or try it right now on doloop's own pages
You can try that same engine right here, pointed at doloop.io instead of your own site. Ask it something clear and it sends you to the right page. Ask something vague and it offers a few options. Ask about something doloop does not cover and it will tell you it cannot answer, rather than making something up. If you are skeptical, try that last kind of question first.
The first three route straight to a page, the fourth is genuinely ambiguous so it returns a menu, the last is declined. Same input, same answer, every time.
Three things it can do, and nothing else
A generated chatbot has an unbounded worst case: it can say anything, and you find out when a customer screenshots it. This has a floor under its worst behavior. The ceiling is lower, it cannot write you a poem or reason through a brand new question, but for a site that cannot afford a public wrong answer, a lower ceiling with a hard floor is the right trade.
Running in the wild
The same engine, seeded from each site's own pages, live on six sites. Two are large business sites with hundreds of pages each, seeded on the full site. The other four each stress one axis of the gate - clean fit, small corpus, hardest corpus, sensitive subject - which is exactly where it earns its keep. Open one and use the launcher in the corner.
Indium enterprise scale
A global AI-services company - hundreds of service and product pages. Seeded on the full site: ask about The Lifter, data engineering, or quality engineering and it routes to the exact page, asks when a question spans several, and refuses what the site does not cover. Scale is the test here, and the suggestions track the section you are reading.
Open indium.tech →iXie Gaming enterprise scale
A game-testing studio - game QA, art outsourcing, porting and remastering across about four hundred pages. The same engine, seeded on its own pages, answers visitors' real questions or declines off-topic ones. The launcher is live in the corner.
Open ixiegaming.com →Doloop Digital clean fit
A dense single-thesis site where every page orbits one idea. It passes both checks: it answers its real questions, and it refuses off-topic ones. The kind of site this is built for.
Open doloopdigital.com →Ekrasworks small corpus
A craft and research brand with distinct essays and a product kit, each answered with its own title, summary, and image. It routes its real questions cleanly, but the vocabulary is small enough that an off-topic question can land near a real page by chance. So it is exactly the site the abstain monitor watches: strong on reachability, the refuse axis is the one to guard, and the live gate flags it for that.
Open ekrasworks.com →Sūtrakṛt Gītā hardest corpus
700 nearly identical verses, about the worst case there is. A verse number lands the exact verse, but a broad concept makes it ask rather than guess. The hard floor on display: it defers often here, and never invents a verse.
Open gita.ekrasworks.com →Hindu Spiritual Care sensitive subject
A nonprofit where a wrong or invented answer would do real harm. So it over-declines on purpose, and that is what keeps it on the safe side of the refuse axis: it routes courses, donating, and chaplaincy, and on anything it is unsure of it says it does not have a page, rather than improvise.
Open hsciglobal.org →It stays current, and stale never means wrong
The widget sits on every page you publish, so it notices when a page changes as visitors load it, then re-reads that page from the source before anything is re-frozen. We only ever freeze content we fetched ourselves, so the answer set always traces to a page we actually read. Until a change catches up, a stale page still routes correctly or falls to a menu or a decline. It is never confidently out of date. A generated chatbot has no such floor: a stale page becomes a fluent wrong answer, with old facts asserted at full confidence and no signal that anything is out of date.
It is for the job most website chat actually is
FAQ, support deflection, navigation, "what is this, where do I, how much, what is your policy." It turns a visitor's question into the right page in one step, instead of a menu hunt. Most of the volume and most of the visitors. It understands paraphrase and typos, costs nothing per answer, and cannot leak or invent.
Transactional jobs (check my balance, cancel my order) and open-ended back and forth. Those need a model that can act and reason. We will tell you when your site is one of these, instead of pretending otherwise.
How it gets onto your site
After you paste your URL at the top and see it work, hit "Put it on your site" and drop your email. We send you the script tag for your domain plus a link to a Google Sheet where you can share more context about your site (optional, but helps us tune it). Drop the tag on every page, the launcher appears in the corner, and it routes to your real pages or refuses honestly. No model runs when it answers, so it costs nothing per question.
The readiness gate you saw at the top - can it answer your real questions, and does it refuse off-topic ones - runs on every build. A site that passes both axes reveals instantly. One that doesn't gets flagged for review before it goes live, so you never ship a bot that guesses.
Want the honest version on one page, including where a generative chatbot still wins? Read the spec sheet.
The same question always gets the same answer, because the bot only ever points to a fixed set of your pages and no model runs when it replies. Ask it something off topic and it hands you off politely. Try to trick it and it refuses. Either way, it never invents an answer.