Grade calls on your own scorecard, not on a feeling
Call QA reads a finished call and grades it on the things you care about. A phrase you banned, a line you require and a step that was missed all show up. A person reviews what the rules pick.
From Professional- Greeting said
- Banned phrase flagged
- Sent to a reviewer
A manager can only hear so many calls
You hear a handful and judge the month
A supervisor listens to a few calls and calls that a review. The call that would have worried you is usually one nobody opened.
Two reviewers, two scores
One supervisor passes what another fails. A person being graded cannot tell which standard they are being held to.
Risky words go unnoticed
A banned promise or a missed notice is one sentence in a long call. Without something reading for it, you find out from the complaint.
From a finished call to a fair grade
Set the standard
Start from a ready-made scorecard or write your own dimensions and pass bar. A Policy pack lists phrases that must be said, phrases that must never be said, and actions that must happen.
Choose what gets graded
Automatic, risky calls only, on demand, or off, in Call Settings. A new workspace starts on demand, so nothing runs or is charged until someone asks.
Read the grade
Each dimension gets a mark. The AI's part of a call and a person's part are graded on their own, so a hand-over never mixes the two.
Review and answer disputes
Your review rules pick which grades a reviewer looks at. A person can dispute a grade once, and hears the answer in their notifications.
Each grade shows you why
A graded call lists the marks by dimension, what the Policy pack caught, and what a reviewer did about it.
- The numbers describe the calls that were graded. The page says how many of how many.
- AI agents and Staff are separate views, so a person's part never lowers an agent's numbers.
- Grades are dated by when the call happened, not when it was graded.
Policy pack
Review
The standard is yours
Scorecards
Your own dimensions and pass bar, for each agent or team. Staff calls can have a scorecard of their own.
Policy pack
Required phrases, banned phrases, required actions and rules. Phrases are matched in what the graded speaker said, never the customer.
What runs by itself
Automatic, Risky calls only (with a share of the rest that you set), On demand, or Off. An agent or team can set its own.
Review rules
They choose which grades a person looks at. They never grade a call themselves.
Calibration
Have several reviewers score the same calls and see where they disagree.
Coaching notes
Press Write coaching notes to draft notes for one person from their graded calls. They read them on their Workspace.
Who opens it, and what they get
QA and floor managers
One standard across reviewers, a queue instead of a random sample, and disputes with a record.
Telecallers
Their own grades and notes on their Workspace, and one chance to dispute a grade they think is wrong.
Whoever owns the script
A clear view of which step agents keep missing, so the fix goes in the right place.
Which plan has it, and what it costs
| Plan | Starter₹999 a month | Growth₹2,999 a month | Professional₹7,999 a month | Enterprise₹28,000 a month |
|---|---|---|---|---|
| Call QA | Not included | Not included | Included | Included |
| Every staff call graded by itself | Not included | Not included | Not included | Included |
What it costs
- ₹0.53 per call graded. Each part of a call graded for the first time is charged once. One hand-over can be two charges: the AI's part and the person's. Grading calls together shows the total first.
- Test calls are never graded. Each Coaching run that writes notes is charged from your wallet at the coach price on your plan.
Call QA is not on the free trial.
See every plan and rateWatch out for this
Set QA grading to Automatic and every answered call is graded and charged. Using a template or resetting to the default replaces that scorecard's dimensions and Policy pack in one step, with no undo.
It works with these
Call QA: questions
Does it grade every call?
Can it grade my people as well as my AI agents?
Can someone challenge a grade?
Is the AI's grade final?
Decide what a good call means. Then check for it.
Start with a ready-made scorecard and edit it. Grading runs only when you ask, until you say otherwise.