Working the Queue

Resonant IQ runs on an auto-approve model. The AI is the primary reviewer: it scores every conversation against your rubric automatically, and that score stands on its own. There's no approval step, no “mark as reviewed,” and no bulk sign-off. Your job isn't to rubber-stamp the AI — it's to catch the scores worth a second look and correct, coach, or flag.

Your QA pulse

The strip across the top of QA answers “where should I look first?” before you touch the queue.

Three states, and only three

A conversation moves through exactly three states. There is no “reviewed” or “complete” — those would imply a human gate that doesn't exist here.

Reading the queue

Four tabs slice the scored conversations by what's actionable. The Unseen only toggle hides anything a manager has already opened, and a ✏ corrected badge marks rows a human has already touched — both visible without opening a single conversation.

The three actions that matter

Open a scored conversation and you'll see the overall score, the per-criterion breakdown, and three actions. Each one writes its own audit trail — that's what separates a real decision from a glance.

Who can do what

Queue actions are gated by role. Agents can dispute their own scores; managers and admins do the correcting and coaching.

What happens next

Coaching notes you add here don't stop at the conversation — they roll into a full coaching workflow with assignments, due dates, and score-improvement tracking. That's covered in Scores, flags & coaching.