Quality monitoring with Yeastar: every call evaluated against your scorecard

AI quality monitoring for 100% of Yeastar PBX calls: a scorecard per queue, DID or extension, scores per employee and evidence-backed feedback per criterion.

BRIGGS Team5 min readLer em português

Short answer

At many companies running a Yeastar P-Series PBX, quality monitoring follows the same routine: the supervisor opens the recordings list, picks a few calls per agent, listens, fills in a spreadsheet and schedules feedback for whenever there's time. The sample is small and depends on who picked it.

With BRIGGS connected to Yeastar, monitoring covers every recorded call in scope. Each one is evaluated against the company's scorecard, with a score, reasoning per criterion and the excerpt from the conversation that backs it up. The supervisor stops listening blind and starts coaching. Integration overview in Yeastar + BRIGGS.


Manual monitoring vs. AI monitoring on Yeastar

Manual With BRIGGS
Calls evaluated a few per agent, picked by hand every recorded call in scope
Supervisor's time hours listening to audio reading the highlights and reviewing whatever they want
Criteria change from one evaluator to the next the same scorecard for everyone
Reasoning a short note, when there is one per criterion, with the excerpt from the call
When feedback happens days or weeks later right after the call is evaluated
Agent disputes hard to check the agent sees what was evaluated and why

The scorecard is your company's

BRIGGS doesn't come with a generic checklist. The company sets up its own scorecard, with the questions it already uses in manual monitoring:

  • "Did the agent introduce themselves and confirm the customer's name?"
  • "Did they explain the timeline and the next step?"
  • "Did they offer the right product for what the customer asked?"
  • "Did they close by checking whether there were any other questions?"

Each question can have its own weight. Knockout criteria, such as failing to give mandatory information, set the score to zero when they occur. The scorecard is edited in the platform, and the next calls are already evaluated against the new version.


A scorecard for each queue, DID or extension

A sales call and a support call can't be judged by the same yardstick. In Yeastar, calls already arrive separated by inbound number (DID), queue and extension. BRIGGS takes advantage of that: queues, DIDs and extensions can each have their own scorecard.

  • The sales number is evaluated on uncovering the need, handling objections and moving toward the close.
  • The support queue, on solving the problem and confirming the fix.
  • The collections queue, on following the mandatory script and negotiating within the rules.
  • The front desk, on understanding the request and routing it to the right place.

That way, each person is held to what they actually do.


Who gets evaluated: the extension owner

BRIGGS uses the extension that answered the call to identify the employee being evaluated. Since each agent in Yeastar usually has their own extension, the score lands on the right person automatically.

That makes it possible to follow each employee over time, criterion by criterion. If last month's training was about presenting the next step, you can see whether that criterion improved in the following weeks.


Calls that don't count

For the score to be fair, it has to look at real conversations with customers. In the integration settings, the company can:

  • skip short calls, such as those that dropped in the first few seconds;
  • skip internal calls between the company's own extensions;
  • set the cutoff date from which recordings are included in the evaluation.

This keeps a five-second call with no conversation from dragging down someone's average.


What the supervisor gets to see

  • Score per employee, by criterion and over time.
  • The criteria that fail most across the whole team. When everyone gets the same point wrong, the problem is usually the process, not the person.
  • Fair comparison, because everyone is evaluated against the same scorecard.
  • At-risk calls, where the customer talked about canceling, complaining or going to Procon (Brazil's consumer protection agency).

Agents can also see their own evaluations, with the excerpt for each criterion. The feedback conversation becomes concrete: "at minute two, the customer asked about the installment amount and never got an answer."


AI evaluates, the supervisor decides

The automatic evaluation is the first step, not the final word. The supervisor can open any call, listen to the audio, read the transcript and adjust the score when they disagree. The change is logged, with who changed what. Over time, those adjustments help calibrate the scorecard.

To understand what the AI analyzes beyond the score, such as contact reasons and sentiment, see speech analytics with Yeastar.


Frequently asked questions

Do I need to change any call-handling settings in Yeastar? No. Queues, extensions and IVR stay as they are. The only thing to allow is BRIGGS's access to the PBX API, which the administrator sets up under Integrations → API.

Can we start with just one team? Yes. The company chooses which queues, DIDs or extensions are included and expands later.

What if the call wasn't recorded? It isn't evaluated, because BRIGGS works from the audio. Recording is configured in Yeastar itself. We explain where recordings are stored and how they reach BRIGGS in call recording on Yeastar.

Does monitoring also work for WhatsApp? Yes. If the team serves customers on WhatsApp, those conversations can come into the same BRIGGS account, with a chat-specific scorecard.

Want to evaluate every call on your Yeastar? Talk to us.