TemplatesGet a demo →Book a meeting

Customer questions answered from your docs, cited, around the clock

Customers ask in their own words and get short, practical steps drawn from your help center and runbooks, each answer cited. When the answer step comes back empty, the question goes to your support team instead of the customer.

Support · Someone asks · 7 steps · Reads your documentsA person is asked at “Ask a person”

AI Factory · First-line supportthe flow as it opens in the builder
QuestionQuestion
RetrieveRetrieve
AnswerAnswer
If / elseGrounded in the sources?
Answer shownAnswer shown
Ask a personAsk a person
Answer shownAnswer shown
7 steps in the flowDrag the canvas, or any step, to move it around.

Why teams run it

A large share of support volume is questions your documentation already answers, but customers do not search it and agents re-type the same steps all day. A generic chatbot answers faster, and also invents steps your product does not have.

How it is governed

  • Every claim must cite your help center or runbooks.
  • Retrieval returns only passages the person asking is permitted to see.
  • Escalated questions wait for your support team’s approval before anything is shown.
  • Every question and answer is kept in the run history.
Who runs it
Support team leads; Customer operations managers; Self-service and knowledge managers.
What you supply
Your help centre or runbooks as a knowledge source.

It ships with test questions

  • How do I reset my password?
  • What are your support hours?
What “a person approves” looks like in the product: the run stops, the banner says what approving will do, and until someone decides, nothing leaves the system. A demo workspace on synthetic data.

From template to a live agent

The same builder, tests and release review apply to a template as to anything built from scratch.

Open it as a working flow
Every step, source, branch and approval is already in place, and every one of them can be changed before it runs on your data.
Prove it before it ships
An evaluation set can gate the release, so a change that breaks an answer never reaches anyone. A question it cannot answer from the sources leaves the gate unproven, not passed.
Keep a person on the decisions
A run that needs a decision stops and says exactly what approving will do, and who it is waiting for.
Publish where people work
A form, an embedded widget, Microsoft Teams or a scheduled batch — with its own access and limits per deployment.
What teams usually change
Make the check stricter, so 'not covered' answers also reach a person; Name the support queue or person who takes escalations; Publish it as a widget, form or link with access rules.
Step 02, in the product: an evaluation set that gates the release. A question it cannot answer from the sources leaves the gate unproven, not passed. A demo workspace on synthetic data.

Questions buyers ask

What does it do when the docs don’t cover a question?
The answer says so plainly rather than guessing, and if the answer step returns nothing at all the question goes to your support team. Many teams tighten the check so that a 'not covered' answer also goes to a person.
Where can customers use it?
Once published, an agent can run behind a public link, a form or a widget on your own sites, with public, signed-in or password access and spending caps. Every run is recorded, refusals included.
Can it be talked into saying something off-script?
It is instructed to answer only from your retrieved passages and to cite every claim. Prompt-injection attempts are scanned for as part of the platform’s protections.

See it run on your own data

We open it in the builder, point it at a sample of your documents and systems, and run it with you.