Skip to main content
Talk to your Account Manager to help get this feature set up for you.
If an AI conversation doesn’t go the way you expect, flag it right from the call. Your Hatch team sees the flag, works it, and closes it out — so you’re not tracking bad calls in a spreadsheet or waiting on an email thread to find out whether anything happened. Flagging is the right tool for problems with the agent itself: a conversational flow that went wrong, poor voice quality, the agent ignoring its instructions, or a scheduler or integration error that stopped a lead being booked.

​
Where you can flag a call

There are three places to raise a flag, and they all do the same thing:
  1. From the Conversations page — click any AI conversation, then flag it from the side drawer.
  2. From the Contact Record — open the Conversations drawer on the contact card and flag any AI conversation there.
  3. While listening to a recording — flag directly from the recording view, whether you opened it from the Contact Record or the Conversations page.

​
Raising the flag

Flagging asks for one thing: a note describing what went wrong. It’s required and free text, so be specific — the note is what your Hatch team uses to diagnose the problem. “Agent offered a Saturday slot we don’t service” is far more actionable than “bad call.” On submit, the conversation is marked Flagged, and the flag is recorded with your name and the time you raised it.
Flags can’t be edited after they’re submitted. If you need to add something, raise a new flag on the same conversation or tell your Hatch account team.

​
Finding your flagged conversations

Flagged is a filter on the Conversations page. Apply it to see only conversations with an active, unresolved flag — which makes it a natural review queue. This works well as a group exercise: flag calls across the week, then sit down with your Hatch team and work through the filtered list together, rather than trading links one at a time.

​
Getting notified about new flags

Flag notifications are currently in beta and need to be switched on for your organization and for each person being notified. Ask your Hatch account team to enable it.
Once enabled, notifications are set up through Automations in the left-hand sidebar:
1

Create a new automation

Go to Automations, then click + New Automation.
2

Set the trigger

Set the trigger to AI Conversation Flagged.
3

Set the action

Set the action to Send Notification, and choose who to notify. Email notifications require the recipient to be a user of that organization.

​
What happens after you flag something

Your Hatch account team picks the flag up, works it, and closes it out. Each flag carries a status through to resolution, so issues don’t disappear into an inbox. A few things worth knowing about how this works today:
  • You aren’t notified automatically when a flag is resolved. Your account team will close the loop with you directly.
  • Resolving flags is currently an internal action. You can’t mark your own flag resolved yet.
  • You can’t tag a specific moment in the transcript. Describe the moment in the note instead.
  • Flags aren’t support tickets. They route to your Hatch account team, not to Support, and there’s no SLA attached.
Flagging is most valuable during onboarding and pilots, when your agent is still being tuned. Reviewing calls and flagging the ones that miss is the fastest way to get your Voice AI agent behaving the way you want.