Chat is the text-based way to reach a requester, available from the communications dock that follows you around the console. It uses the same AI end-user model that voice calls and NorthMail-linked mail threads use — a message you send in chat and a call you place five minutes later about the same ticket are, from the requester's side, one continuous conversation with the same person, not two disconnected interactions. This is deliberate: a real requester does not reset their memory every time you switch from typing to talking, and neither does the simulation.
Sending a message
- Open the communications dock and select the requester tied to your active ticket, or open chat directly from the ticket workspace.
- Type your message. The character counter shows how close you are to the limit as you type.
- Send it. The requester's reply appears in the thread within a few moments, generated live rather than pulled from a script.
- Continue the thread for as long as the incident needs — the whole conversation stays attached to that ticket and is visible to you if you come back to it later.
How the composer works
- Messages are limited to 500 characters, so say what you actually mean rather than pasting a wall of text — a real chat message wouldn't be a paragraph either.
- There is a minimum length on a sent message, so a message too short to actually be a real update won't send.
- Sends are rate-limited — you can't send messages back-to-back with no gap, which keeps the pacing closer to a real conversation instead of letting you flood the thread.
Chat is tied to a specific ticket the same way mail and calls are. When you open a chat thread from a ticket's workspace, the requester on the other end responds with the context of that specific incident, and anything you or they said earlier in that ticket's history — across chat, call, and mail — is available to them. This is what makes a follow-up like "hi again, this is about the printer thing" land correctly instead of the AI acting like it's meeting you for the first time.
As with calls, an employee can message you before any ticket exists — an unprompted chat about a problem you haven't seen yet, which creates the ticket the moment you respond to it. Free accounts share 25 AI-assisted chat and mail messages per day. Both paid plans include unlimited chat and mail responses.
Chat responses, like call responses, are graded as part of a ticket's communication score. Sending nothing, sending something too vague to move the ticket forward, or skipping identity verification on a request that called for it all show up in how that ticket scores, not just whether it eventually closes.
Chat is generally the faster and lower-friction of the two live communication channels, since there's no microphone permission to grant and no live audio to manage — it works identically on a laptop with no working mic or in a browser tab where you'd rather not talk out loud. Many technicians default to chat for quick clarifying questions and reach for a call only when the ticket genuinely benefits from a real-time back-and-forth, the same trade-off a real support team makes between a quick message and picking up the phone.
Every chat thread stays attached to the ticket it started on, visible from that ticket's workspace for as long as the ticket exists. If a ticket reopens later, the earlier thread is still there — the requester on the other end still remembers it, and so do you, since the full history is right where you left it rather than something you have to reconstruct from a separate messaging log.