Quality command center

See how the team is performing this week and where coaching is needed.

Calls Reviewed01

0

Reviewed calls in selected week

Quality ScoreQA

0%

Weighted team QA average

Booking RateBK

0%

Calls saved as booked

Call EfficiencyWO

0%

No work orders reviewed yet

Agent performance

Weekly ranking by weighted quality score

Team quality categories

Average score by QA dimension

Customer Data First

MP3 → datos críticos → evidencia → verificación QA → historial. Sin puntuaciones automáticas.

1. Grabación

Archivo MP3 o WAV

Se envía temporalmente a Deepgram; no se almacena audio en D1/R2.

2. Referencias del trabajo (opcionales)

Job #

Agente

Notas del agente / Work Order

Carga una llamada para comenzar.

Historial · Customer Data

Revisiones verificadas y pendientes. No almacena archivos MP3.

Selecciona Historial Data para cargar tus revisiones.

Evidence-first QA workspace

Review a call

Listen, verify, and document what actually happened. Work Order is optional.

01 · Call & context02 · Optional Work Order03 · Evidence review04 · QA approval

1Call informationOnly what QA needs to enter

Agent

Incoming number

Website

Service, city, call intent, booking result, cancellation reason and call date are detected automatically from the recording. City is taken only from the service address stated in the call.

2Call recordingTemporary processing only

A recording is required. Audio is temporary and the full transcript is not kept with saved QA reports. Work Order notes and screenshots are optional.

♪

Call recording

MP3, M4A, WAV, OGG, WEBM or MP4. Up to 20 MB.

3Work Order / agent notesOptional — no document means no documentation penalty

For a booking, paste the agent's notes or upload a screenshot. For not-booked calls, cancellations, reschedules or follow-ups, select No Work Order. The audio can always be evaluated by itself.

No Work Order Paste agent notes Upload screenshot

Call-only QA: documentation accuracy is N/A. No points are deducted for missing Work Order evidence.

Call details

Critical customer and booking data appears first. QA scoring loads later at the bottom.

Mandatory human verification

Final call classification & fair scoring

Actual result of this call

AI proposed classification

Waiting for analysis

A lost booking is not a QA failure by itself. Extra-services and lead-source questions apply to sales/new appointments, not administrative, cancellation or follow-up calls without a new booking.

I reviewed the evidence, N/A criteria and proposed scores. No penalty is based solely on unclear audio, missing notes, or AI guesses.

Customer & appointment detailsVerified / Partial / Not provided

Specific service

Service category

Customer name

Email

Primary phone number

Alternate phone number

Alternate-number question

Appointment day / date

Appointment time / window

Booking outcome

2-hour window compliance

Visit / inspection fee

Actual service price

Previous estimate referenced

Service price accepted?

Tip / gratuity

Lead source

Required lead-source question

Additional services / upsell

Full service address

Address double-checkClick each timestamp to hear that exact part

Critical fields are shown before scoring. You can correct any value manually; the timestamp remains the original speech evidence.

Call intent

Not analyzed yet.

What happened in this call

The call story will appear here after analysis.

Coaching observations — QA must verify0 issues detected

What the agent didObjective checklist from the full call

Agent Timeline — every agent utteranceScroll inside this panel

Possible missed opportunitiesAI suggestions — QA must verify before coaching

Important moments in the callClick a timestamp to jump in the recording

QA Notes & follow-upManual notes written by the QA reviewer

QA notes

Feedback sent to agent?

Mark Yes only after QA feedback has actually been delivered.

Verified Flags are rule-based. Possible Misses are suggestions and should be confirmed by QA.

QA percentagesRule-based · calculated instantly after transcription

0Overall QA

Weighted score based only on applicable categories.

Call Efficiency will appear after analysis.

CORE scoring uses deterministic transcript/timestamp rules. Tone, friendliness and listening are transcript-based behavioral proxies; objective items such as customer information, lead source, booking actions, dead air and upsell use direct evidence.

QA Queue

Live activity for Denisse, Marla, Oscar and Charly · integrated directly into QA Intelligence.

QA Online

0/4

Active, available or reviewing

Reviewing

0

Calls currently being checked

Available

0

Inside QA Intelligence, not in a review

Completed Today

0

Saved QA reviews today

Live QA activity

Status is based on real QA Intelligence activity, not just mouse movement.

—

Loading QA Queue…

Reports

Operational trends from saved QA records, separated so each report loads only when you open it.

Reports reflect saved QA review records. A Viewer account only receives its linked agent's data.

Weekly performance

Compare agents across booking, overall quality, customer-data accuracy, and individual QA categories.

Team QA

0%

Booking Rate

0%

Calls Reviewed

0

Call Efficiency

0%

Agent scorecard

Percentages reflect saved reviews for the selected week.

Call history

Review saved QA records and reopen detailed feedback.

Manager settings

Manage agents, privacy, AI status and system connections.

Agents

Names available in reviews and weekly scorecards.

Privacy & AI usage

Audio is temporary; only historical performance data is retained.

Privacy rule: audio is deleted after processing, the full transcript is not saved, and customer phone/email/address are not kept in the historical database.

System

Cloudflare, Deepgram and Workers AI configuration status.

Manager permissions

Your manager role has expanded controls.

Manager: review calls, edit agent roster, view all history, delete QA reviews, manage users, change retention settings, run cleanup and open the database.

QA agent: review calls and see performance/history, but cannot change administration settings.

Viewer: read-only dashboard, weekly performance and history.

Users & access

Create staff logins and control what each person can change.

Call quality with evidence, not guesswork.

Sign in

Create manager account

Quality command center

Agent performance

Team quality categories

Customer Data First

1. Grabación

2. Referencias del trabajo (opcionales)

3. Datos para confirmar

4. Resultado operativo (sin calificación)

Comparación con Work Order (solo diferencias candidatas)

Historial · Customer Data

Review a call

Documentation accuracy · evidence comparison

Call details

Final call classification & fair scoring

QA Queue

Live QA activity

Reports

Weekly performance

Agent scorecard

Call history

Manager settings

Agents

Privacy & AI usage

System

Manager permissions

Users & access

New user