What it does
Every reply työ drafts is held until a person approves it. You read the draft, change anything you want — or rewrite it completely — and send. Only then does the message reach the customer, in their language. There is no setting that turns this off, because governed messaging is what työ is.
For replies that carry more weight — a refund, a complaint, a promise about price or timing — you can require sign-off from more than one person. You decide which kinds of replies need a second pair of eyes, and the work won't go out until everyone required has approved it.
Why it matters
Your name is on every message. When a customer asks whether the kitchen can handle a nut allergy, or whether a refund is coming, the answer is a commitment your business has to keep. A chatbot will happily improvise that commitment for you; työ won't. A person who knows your business makes the call, every time.
Customers can tell the difference. A reply that a human stood behind builds the kind of trust a small business runs on — and it protects you, too, because nothing goes out that you didn't see. It also puts you on the right side of where regulation is heading: the EU AI Act rewards exactly this kind of human oversight.
How it works
When a customer message arrives, työ routes it to the right team member and prepares a draft grounded in your business's own knowledge. The draft sits in review — ready in seconds, but going nowhere on its own.
The reviewer opens it, edits, and approves. If a reply is marked sensitive, työ collects the configured number of sign-offs before sending. And if a draft just sits there, it doesn't quietly expire: the work escalates to a teammate and resurfaces in your daily briefing until someone closes it.