Dashboards & Analytics · Call Outcomes

Call Outcome Dashboard for Voice Agents

A reporting dashboard that summarizes voice-agent call outcomes — connected, qualified, booked, escalated — so the operator can read agent performance without scrubbing transcripts.

Start a ProjectAll Case Studies
5 to 8 days
build time
4
outcomes
5
stack tools
6
build steps

Built with real HMX dashboard tool paths

Supabase PostgresCall-outcome records (existing writes)SQL viewNext.js 16 server componentsMetabase (read-only BI)Supabase PostgresCall-outcome records (existing writes)SQL viewNext.js 16 server componentsMetabase (read-only BI)

01 // Outcomes

Outcome signals

Outcome
breakdown for connected, qualified, booked
Escalation
share to human surfaced clearly
Rate
trends without scrubbing transcripts
Flagged
calls with missing outcome data

Case architecture

Call Outcome Dashboard for Voice Architecture

6 nodes
the call-outcome fields
Agree the outcome categories
Supabase Postgres
Call-outcome records
Review Queue
Dashboard Action
  1. 01the call-outcome fields

    A reporting dashboard that summarizes voice-agent call outcomes — connected, qualified, booked, escalated — so the operator can read agent performa...

  2. 02Agree the outcome categories

    Agree the outcome categories and rate definitions (qualified rate, booked rate) with the operator.

  3. 03Supabase Postgres

    Supabase Postgres contributes the trusted model for Call Outcome Dashboard for Voice so metrics are defined before they are visualized.

  4. 04Call-outcome records

    Build a view summarizing outcome counts, rates, and escalation share over the window.

  5. 05Review Queue

    When automation confidence is low, route the record to a manual owner with the source, stage, and last action attached.

  6. 06Dashboard Action

    Outcome breakdown for connected, qualified, booked; Escalation share to human surfaced clearly; Rate trends without scrubbing transcripts; Flagged...

Problem

The operating gap

Voice-agent results sit in call logs and transcripts; answering 'how many calls qualified or booked this week, and how many got escalated to a human' means reading through records one by one, so the outcome picture is never current.

Build

What gets built

A read-only dashboard over stored call-outcome records that summarizes outcome categories, qualification and booking rates, and escalation counts over a window, with a recent-calls outcome list. It is strictly a reporting surface over outcome data the agent already writes — it does not run or change the agent.

Build steps

How it ships

Call Outcome Dashboard for Voice Agents uses a reporting model and review layer for Dashboards. A reporting dashboard that summarizes voice-agent call outcomes — connected, qualified, booked, escalated — so the operator can read agent performa... The architecture connects the call-outcome fields, supabase postgres, call-outcome records, and dashboard action with an explicit control path.

  1. 01Confirm the call-outcome fields recorded (connected, qualified, booked, escalated, duration) and their values.
  2. 02Agree the outcome categories and rate definitions (qualified rate, booked rate) with the operator.
  3. 03Build a view summarizing outcome counts, rates, and escalation share over the window.
  4. 04Render a dashboard with outcome breakdown, rate trends, and a recent-calls outcome list.
  5. 05Flag calls with missing or ambiguous outcome data so they do not distort the rates.
  6. 06Document the categories so call-outcome reporting stays consistent across weeks.

Stack

Tools and layers

  • Supabase Postgres
  • Call-outcome records (existing writes)
  • SQL view
  • Next.js 16 server components
  • Metabase (read-only BI)
  • Inputs layer: Confirm the call-outcome fields recorded (connected, qualified, booked, escalated, duration) and their values.
  • Transform layer: Agree the outcome categories and rate definitions (qualified rate, booked rate) with the operator.
  • Metrics layer: Supabase Postgres contributes the trusted model for Call Outcome Dashboard for Voice so metrics are defined before they are visualized.
  • Visualization layer: Call-outcome records (existing writes) handles refresh, review, or reporting delivery while a read-only dashboard over stored call-outcome records that summarizes outcome categories, qualification and booking rates, and escalation counts o...
  • Action layer: Outcome breakdown for connected, qualified, booked; Escalation share to human surfaced clearly; Rate trends without scrubbing transcripts; Flagged...

Data flow

  1. 01Confirm the call-outcome fields recorded (connected, qualified, booked, escalated, duration) and their values.
  2. 02Agree the outcome categories and rate definitions (qualified rate, booked rate) with the operator.
  3. 03Build a view summarizing outcome counts, rates, and escalation share over the window.
  4. 04Render a dashboard with outcome breakdown, rate trends, and a recent-calls outcome list.
  5. 05Flag calls with missing or ambiguous outcome data so they do not distort the rates.
  6. 06Document the categories so call-outcome reporting stays consistent across weeks.

Controls

  • Voice-agent results sit in call logs and transcripts; answering 'how many calls qualified or booked this week, and how many got escalated to a huma...
  • A read-only dashboard over stored call-outcome records that summarizes outcome categories, qualification and booking rates, and escalation counts o...
  • When automation confidence is low, route the record to a manual owner with the source, stage, and last action attached.

Research basis