UMAR Case study · SCO

How SCO reviews around ten times as many calls with UMAR.

Special Communications Organization (SCO) had two QA managers reviewing calls by hand, so only a small sample was ever heard. We delivered UMAR to transcribe recorded calls, score them against SCO's QA criteria and put the results in front of managers, who still review the scores.

See UMAR Book a demo

Completed for SCO. About 10× QA throughput at a similar operating cost.

Call review view. Illustrative data.

Client
Special Communications Organization (SCO)
Status
Completed
Result
About 10× QA throughput at a similar operating cost
Scoring
SCO's own QA criteria

The problem

Two QA managers could only hear a small sample of calls.

QA depended on people listening and scoring by hand, so coverage was limited by the hours the two managers had.

  • Most calls went unreviewed

    QA covered only a small share of the calls SCO handled.

  • Consistency depended on time

    Scores relied on how much time and attention each review could get.

  • Patterns were hard to see

    With a small sample, recurring issues across the operation were difficult to spot.

  • More coverage meant more cost

    Reviewing more calls by hand would have meant more people and more pressure.

What SCO gained

Cheaper, more consistent QA, with managers still in charge.

Ten times the throughput was the headline. It also changed what QA could see and how fairly calls were scored.

  • 10×

    Far more calls reviewed

    Around ten times the QA throughput from the same team.

  • Similar cost

    Cheaper per reviewed call

    Throughput rose without a matching rise in operating cost, so each reviewed call costs far less.

  • Same rubric

    Consistent scoring

    Every call is assessed against the same structured QA criteria.

  • Visible

    Patterns come to light

    The dashboard shows reviewed calls, scores and recurring QA themes across the operation.

  • Flagged

    Attention where it is needed

    Calls that need a closer look are surfaced for manager review.

  • In charge

    People stay in control

    QA managers review the results before they are used.

What we built

From the recording to a score managers can act on.

We covered the whole path, from collecting the calls to a reporting dashboard for SCO's QA managers.

  • Call intake

    A repeatable way to bring call recordings into review.

  • Transcription

    Speech-to-text turns each call into text that can be scored.

  • QA scorecard

    SCO's QA criteria set out so every call is judged against the same standard.

  • AI scoring

    Large language models review each transcript against the scorecard.

  • Structured scores

    Scores laid out for QA managers to compare and follow up.

  • Review and exceptions

    Managers check the results and deal with calls that need a closer look.

  • QA dashboard

    Reviewed calls, scores and recurring themes in one reporting view.

  • Completed and delivered

    The finished UMAR solution, delivered for SCO's call QA process.

How it works

Five steps from a recorded call to the dashboard.

The same path for every call, with people checking the results.

  1. Calls are collected

    Recordings enter a repeatable review process.

  2. Speech becomes text

    Each call is transcribed automatically.

  3. AI scores the call

    The transcript is checked against every criterion on the scorecard.

  4. Managers review

    QA managers check the scores and deal with flagged calls.

  5. Results reach the dashboard

    Reviewed calls and recurring themes are visible in one place.

Your scorecard, applied the same way every time

Consistent scoring, with people still in charge.

UMAR works from defined QA criteria, not open-ended analysis. It applies the same scorecard to far more calls and passes the results to SCO's QA managers.

  • Clear criteria. Quality standards written as criteria that can be applied the same way every time.
  • Automated scoring. Transcripts are reviewed at scale against the agreed scorecard.
  • Manager oversight. People review the results before they are used.
  • Exceptions flagged. Calls that need closer attention are surfaced for another look.

Call review view. Illustrative data.

10×QA throughput

What it delivered for SCO.

UMAR gave SCO around ten times the QA throughput at a similar operating cost. Managers can now see a much larger share of calls, scored against the same criteria, with human review still in place.

See UMAR

Bring a sample of your calls. We will show you how UMAR scores them.

A demo takes about 30 minutes and is led by the team that built UMAR.

Book a demo See UMAR
  • For contact centre, QA and customer service managers
  • Cloud or on-premise, including air-gapped networks
  • Need something similar built? See our AI services