How it works

Signing up starts a free trial. There's no card required, and the trial is when we build your assistant — so you're evaluating one built on your actual policies, not a generic demo. The 14 days start the day it goes live, not while we build it.

Your dashboard shows how many days remain, and we email you before it ends. If the trial finishes without a plan, your assistant pauses — nothing is deleted, and everything picks up exactly where it left off when you subscribe.

This is the only real work on your side. Send whatever already exists:

• Your employee handbook and HR policies — leave, conduct, remote work, expenses, IT use, health and safety, benefits. PDF, Word or plain text are all fine.
• Your onboarding checklists — what a new hire needs in their first days and weeks, and who is responsible for each item.
• Your org chart and who-to-ask contacts: who handles payroll queries, IT, facilities, and who in HR should receive sensitive matters.
• Your leave types, working week and public holidays, so leave is counted the way you count it.
• The questions your people actually ask, especially the ones that aren't written down anywhere yet.

Your employee list is loaded through a secure import from a spreadsheet or your HR system — please don't email it to us or paste it into forms. It does not need to be tidy or complete. Structuring it is our job, and we'll come back to you with specific questions where something is missing.

Our setup team works inside your account and does all of the following for you:

• Builds your knowledge base from what you sent, so the assistant answers from your policies rather than from general internet knowledge — and tags what managers only should see.
• Builds your onboarding templates: the tasks, who owns them, and when each is due relative to the start date.
• Sets up your leave types and who approves, your policy list, and who each kind of sensitive matter is routed to.
• Writes the assistant's instructions — its tone, its limits, and when it should bring in a person.
• Connects your channels and tools, such as your intranet page, WhatsApp or Telegram, calendar and IT ticketing.

Every action we take on your behalf is written to your audit log with who did it, when, and what content it was based on. Our team building your assistant does not need — and by default does not get — access to confidential HR cases or to the documents your employees upload.

When it's ready, we hand it to you to test. Talk to it the way your people would, and push on the awkward cases: a question your policies don't cover, a request for someone else's pay, a new hire asking what to do first, and a worried message that should go straight to a person.

Before it can go live we also check the essentials are in place: someone named to receive sensitive matters, crisis and support contacts, a consistent org chart, and a working identity check. Tell us what's wrong and we adjust it.

The assistant cannot go live until you explicitly approve it. That is a hard rule, not a courtesy — we build and test, you sign off. Until you do, it never speaks to one of your employees.

Once approved, we put it where your people are:

• Your intranet or careers page — one line of code, which we'll give you (or hand to whoever manages your site).
• A shareable link, for offer packs, email signatures or a QR code on the wall, with no site changes at all.
• WhatsApp and Telegram — the same assistant, the same knowledge, answering everywhere.

Your dashboard is where you see what is happening:

• Onboarding — every new hire, their progress, and what is overdue. Mark tasks done, unblock them, or request a document.
• People — your employee directory, the list the assistant checks when it verifies who it is talking to.
• Leave — requests the assistant submitted, waiting for a person to approve or decline. The assistant never approves anything itself.
• Cases — confidential matters passed to HR. Only the owner and people the owner assigns can open these.
• Approvals and the Audit Log — anything held for a human decision, and every change made in your account.

You don't have to act on any of it. We review the same reports and update your knowledge base as your policies change — that upkeep is part of the service, not something we hand back to you.

• It answers from your documents and says so when it can't — it never invents a policy or an entitlement.
• It only discusses a person's own records, after a code sent to their work email on file confirms who they are.
• It never approves leave, judges anyone's performance or conduct, or gives legal, medical or tax advice.
• It never asks people to type ID numbers, bank details or medical information into the chat — it sends a private upload link instead.
• When something sensitive comes up, it responds with care, opens a confidential case for your HR team, and steps back.

It supports your HR team. It does not replace them.

Raise it from the Support section of your dashboard and it goes straight to our team with your account context attached. If the assistant said something it shouldn't have, tell us — we'll fix the underlying knowledge or instruction, not just the one answer.