Call detail (two-pane page)

Call detail is where you debug with evidence.

The page is built as a two-pane workflow:

• Left: recording + transcript workspace
• Right: Call Insights rail (tabs)

At a glance

• Use transcript search to jump to relevant moments.
• Use Details/Webhooks/Cost/Analysis for operational understanding.
• Some tabs are role-gated (Tools/Latency/Diagnostics/Telephony/Raw).

Left pane: recording and transcript

Recording

If recording is present, you’ll see playback controls (play/pause, seek, mute, duration) and open/download actions.

If recording is missing, the UI explicitly shows “Recording not available.”

Transcript and search

Use the search box to search transcript, speaker, and tool markers.

Common states:

• No transcript available
• No matches found

Right rail: Call Insights tabs

Tabs include:

• Details
• Webhooks
• Cost
• Analysis

Internal roles may also see:

• Tool calls
• Latency
• Diagnostics
• Telephony
• Raw data

If there are too many tabs, extra tabs move under More.

Role and plan differences (important)

• Some debug tabs require internal platform permissions (Tools/Latency/Diagnostics/Telephony/Raw).
• Some list columns and filters on the Call Logs list page are plan-gated (for example: contact/campaign context).

If you’re documenting this for customer operators, be explicit about which tabs they will not see.

Details tab (what it contains)

Details is the operator baseline. It typically contains sections like:

• Call details (timestamps, direction, provider, status)
• Participants (from/to identities)
• Configuration (agent/campaign configuration)
• Summary (when present)
• Outcome or Campaign context (when linked to a campaign task)
• Follow-up (when linked)
• Extracted data (structured fields)

It can also include a relink flow when the session is not linked to the right contact.

Webhooks tab (integration evidence)

Use Webhooks when something “should have logged” but didn’t.

Expect sections like:

• pre-call API evidence
• post-call webhook evidence and delivery results

Cost tab (billing evidence)

Use Cost when investigating spend anomalies.

Expect:

• the customer-visible cost block
• internal-only cost fields for privileged roles

Analysis tab (structured insights)

Use Analysis when you need structured extraction beyond transcript reading.

Common actions:

• run analysis (select a prompt)
• inspect per-prompt results

Tool calls tab (internal only)

Tool calls is for investigating action traces: what tools were called, with what input/output.

It typically supports:

• filtering for success vs failure
• viewing input/output payloads

Latency tab (internal only)

Latency is for diagnosing “why the call felt slow.”

It can include:

• voice-to-voice latency
• STT/LLM/TTS breakdown
• queue-to-connect time
• turn-by-turn tables

Diagnostics tab (superadmin only)

Diagnostics is for deep runtime inspection and linkage.

Expect:

• identifiers
• runtime variables
• errors
• linkage and internal trace metadata

Telephony tab (internal only)

Telephony is for provider-level debug evidence.

Expect:

• answering machine detection
• webhook events
• outbound call logs
• normalized fields and request/response payload blocks

Raw data tab (role-gated)

Raw data provides a JSON payload view.

The payload scope can differ by role (customer-sanitized vs internal-full).

Recommended operator flow

1. Verify recording and transcript exist.
2. Read Details for context.
3. Check Webhooks for integration failures.
4. Check Cost for billing anomalies.
5. Use Analysis when you need structured insights.