Conversations now tell you why the AI did not reply
When the platform decides not to draft a reply, it now says so in the thread, in plain language: mailing-list headers, an automatic sender, a no-reply address, a reply limit reached, a message judged outside your scope, no agent on the queue. Until now that reason existed only in our server logs, so a thread with an inbound and no draft looked the same whatever the cause.
These notes appear under the message they are about — in the email thread, and in the Conversation Details pane on chat, Meta and voice. They are written in your workspace language.
You can write notes too. Use them the way you would in any helpdesk: “called them back, waiting on the PO number.” Your team sees them; the customer never does, and the AI does not read them. You can edit and delete your own notes; the ones the platform writes stay as a record.
Last reviewed . Tell us when a step no longer matches the product.