---
title: "Retrieve Call Log"
method: GET
path: "/v1/retrieve-call-log/{call_id}"
tags: ["Call Logs"]
---

# Retrieve Call Log

`GET /v1/retrieve-call-log/{call_id}`

Retrieve call logs for a specific agent given the agent ID.
Returns a list of call objects, each with an additional "evaluations" field
containing a list of evaluations joined via call_id.

## Path parameters

- `call_id` string, required

## Headers

- `X-API-Key` string, required

## Response `200`

Successful Response

- RetrieveCallLogResponse
  - `call_log` CallLog, required
    - `id` string, required — Unique identifier for the call
    - `agent_id` string, required — ID of the agent associated with the call
    - `external_agent_id` string, nullable — External ID of the agent associated with the call
    - `recording_url` string, nullable, required — URL to the recording file (supported formats: .mp4, .wav, .m4a)
    - `start_time_utc` string, required — When the call started, in UTC
    - `participants` ParticipantResponse[], required — List of call participants
      - `role` 'AGENT' | 'USER' | 'DIGITAL_HUMAN' | 'agent' | 'digital_human', required — Role of a participant in a call or simulated conversation.
      - `spoke_first` boolean, nullable — Whether the participant spoke first in the call
      - `name` string, nullable — Name of the participant
      - `phone_number` string, nullable — Phone number of the participant (if applicable)
    - `call_direction` 'INBOUND' | 'OUTBOUND' — Enum representing the direction of a call.
    - `interface` 'PHONE' | 'WEB' — Enum representing the interface used for a call.
    - `status` 'INITIALIZING' | 'QUEUED' | 'EVALUATING' | 'COMPLETED' | 'INCOMPLETED', required — Enum representing the possible states of a call.
    - `created_at` string, date-time — When this request was created (defaults to now in UTC)
    - `transcript_url` string, nullable — URL to the transcript file (supported formats: .txt, .json)
    - `tool_calls` ToolCallData[], nullable — List of tool calls made during the call
      - `name` string, required
      - `start_offset_ms` integer, required
      - `description` string, nullable
      - `parameters` object, nullable
      - `output` unknown
    - `events` Event[], nullable — List of events that occurred during the call
      - `title` string, required — Title of the event
      - `start_offset_ms` integer, required — Start offset of the event in milliseconds
      - `description` string, nullable — Description of the event
      - `end_offset_ms` integer, nullable — End offset of the event in milliseconds
      - `tags` string[], nullable — List of tags associated with the event
      - `metadata` object, nullable — Additional metadata associated with the event
    - `tags` string[], nullable — List of tags associated with the call
    - `duration_ms` integer, nullable — Duration of the call in milliseconds
    - `evaluations` Evaluation[], nullable — List of evaluations associated with the call
      - `id` string — Unique identifier for the evaluation result
      - `created_at` string, date-time — When this evaluation result was created (defaults to now in UTC)
      - `hallucination` boolean, nullable — Whether the agent exhibited hallucination during the call
      - `hallucination_reasoning` string, nullable — Reasoning behind the hallucination, if applicable
      - `redundancy` boolean, nullable — Whether the agent exhibited redundancy during the call
      - `redundancy_reasoning` string, nullable — Reasoning behind the redundancy, if applicable
      - `agent_speak_percentage` number, nullable, required — Percentage of the call where the agent was speaking
      - `avg_agent_latency` number, nullable, required — Average latency of the agent's responses during the call
      - `latency` Latency
        - `avg_agent_latency` number, nullable — Average latency
        - `max_agent_latency` number, nullable — Maximum latency
        - `p50_agent_latency` number, nullable — 50th percentile latency
        - `p90_agent_latency` number, nullable — 90th percentile latency
        - `p95_agent_latency` number, nullable — 95th percentile latency
        - `p99_agent_latency` number, nullable — 99th percentile latency
      - `filler_ignored_latency` FillerIgnoredLatency
        - `avg_filler_ignored_latency` number, nullable — Average filler ignored latency
        - `p50_filler_ignored_latency` number, nullable — 50th percentile filler ignored latency
        - `p90_filler_ignored_latency` number, nullable — 90th percentile filler ignored latency
        - `p95_filler_ignored_latency` number, nullable — 95th percentile filler ignored latency
        - `p99_filler_ignored_latency` number, nullable — 99th percentile filler ignored latency
      - `num_turns` integer, nullable — Total number of turns taken by the agent during the call
      - `call_summary` string, nullable — Summary of the call, if applicable
      - `custom_evals` union — List of custom evaluations performed during the call
        - unknown[]
          - unknown
        - object
      - `custom_evals_success_rate` number, nullable — Success rate of the custom evaluations performed during the call
      - `goal_success` boolean, nullable — Whether the goal was successfully achieved during the call
      - `goal_reasoning` string, nullable — Reasoning behind the goal success, if applicable
      - `pronunciation_score` number, nullable — Pronunciation score of the agent during the call
      - `pronunciation_reasoning` string, nullable — Reasoning behind the pronunciation score, if applicable
      - `sentiment_label` string, nullable — Overall sentiment label for customer utterances
      - `sentiment_score` number, nullable — Overall sentiment score for customer utterances
      - `sentiment_scores` object, nullable — Per-label sentiment scores for customer utterances
      - `custom_metrics` CustomMetric[], nullable — Custom metrics associated with the evaluation
        - `name` string, required — The name of the custom metric
        - `description` string, required — The description of the custom metric
        - `response_type` string, required — The response type of the custom metric
        - `response_value` string, required — The response value of the custom metric
        - `reasoning` string, nullable — The reasoning of the custom metric
        - `min_value` number, nullable — The minimum value of the custom metric
        - `max_value` number, nullable — The maximum value of the custom metric
        - `category` string, nullable — The category of the custom metric
        - `tags` string[], nullable — The tags of the custom metric
        - `scoring_guidance` string, nullable — The scoring guidance of the custom metric
        - `enum_options` string[], nullable — The enum options of the custom metric
        - `human_edited` boolean, nullable — Whether this metric result was manually overridden by a human
        - `human_response_value` string, nullable — Human-overridden response_value; null when no override applied
        - `human_reasoning` string, nullable — Human-overridden reasoning; null when no override applied
        - `override_reason_code` string, nullable — Structured reason for the latest human override. One of: ai_hallucination, overly_strict_success_criteria, overly_strict_system_prompt, other.
        - `override_notes` string, nullable — Freeform notes from the human reviewer accompanying the latest override
        - `last_edited_by` string, nullable — User id of the most recent human reviewer
        - `last_edited_at` string, date-time, nullable — Timestamp of the most recent human override
      - `human_edited` boolean, nullable — Whether any result in this evaluation was manually overridden by a human
      - `human_goal_success` boolean, nullable — Human-overridden goal_success; null when no override applied
      - `human_goal_reasoning` string, nullable — Human-overridden goal_reasoning; null when no override applied
      - `human_hallucination` boolean, nullable — Human-overridden hallucination verdict; null when no override applied
      - `human_hallucination_reasoning` string, nullable — Human-overridden hallucination reasoning; null when no override applied
      - `human_redundancy` boolean, nullable — Human-overridden redundancy verdict; null when no override applied
      - `human_redundancy_reasoning` string, nullable — Human-overridden redundancy reasoning; null when no override applied
      - `human_pronunciation_score` number, nullable — Human-overridden pronunciation score; null when no override applied
      - `human_pronunciation_reasoning` string, nullable — Human-overridden pronunciation reasoning; null when no override applied
      - `override_goal_reason_code` string, nullable — Reason for the human override of goal_success.
      - `override_goal_notes` string, nullable — Freeform notes from the reviewer for the goal override.
      - `override_hallucination_reason_code` string, nullable — Reason for the human override of hallucination.
      - `override_hallucination_notes` string, nullable — Freeform notes from the reviewer for the hallucination override.
      - `override_redundancy_reason_code` string, nullable — Reason for the human override of redundancy.
      - `override_redundancy_notes` string, nullable — Freeform notes from the reviewer for the redundancy override.
      - `override_pronunciation_reason_code` string, nullable — Reason for the human override of pronunciation_score.
      - `override_pronunciation_notes` string, nullable — Freeform notes from the reviewer for the pronunciation override.
      - `last_edited_by` string, nullable — User id of the most recent human reviewer
      - `last_edited_at` string, date-time, nullable — Timestamp of the most recent human override
      - `predicted_age` number, nullable — Estimated age of the caller in years, from their audio
      - `customer_gender` string, nullable — Predicted gender of the caller: 'male', 'female', or 'indeterminate' when confidence is at or below 0.7
      - `gender_confidence` number, nullable — Confidence 0-1 behind customer_gender; null when the prediction is 'indeterminate'
      - `detected_language` string, nullable — Language detected on the call
      - `language_confidence` number, nullable — Confidence 0-1 behind detected_language
      - `agent_audio_clarity` number, nullable — Agent audio: audio clarity score 0-100 (higher better)
      - `user_audio_clarity` number, nullable — User audio: audio clarity score 0-100 (higher better)
    - `metrics` MetricResult[], nullable — Metric results scored on this call. Includes the audio measurements that have no evaluations column, such as user_perceived_loudness, user_audio_clipping and user_audio_dropouts
      - `metric_name` string, required — Stable name of the metric, e.g. 'user_perceived_loudness'
      - `metric_type` string, required — One of int, float, boolean, enum, qualitative, json, pass_fail, yes_no
      - `value` string, required — The effective value as a string: the human correction when one was applied, otherwise the model's
      - `human_edited` boolean — Whether a reviewer overrode this result
      - `reasoning` string, nullable — Reasoning behind the effective value
    - `annotations` unknown[], nullable — Annotations associated with the call
      - unknown
    - `metadata` object, nullable — Additional metadata associated with the call
    - `custom_metrics` unknown[], nullable — Please use the custom metrics in the evaluations instead
      - unknown
    - `conversation_ended_by` 'AGENT' | 'USER' | 'DIGITAL_HUMAN' | 'agent' | 'digital_human' — Role of a participant in a call or simulated conversation.
  - `status` string — Status of the response

## Other responses

- `422` — Validation Error

## Changes

- **2026-08-28** `bf69ab7268a4` — 6 info
  - added the optional property `call_log/evaluations/anyOf[subschema #1]/items/customer_gender` to the response with the `200` status
  - added the optional property `call_log/evaluations/anyOf[subschema #1]/items/detected_language` to the response with the `200` status
  - added the optional property `call_log/evaluations/anyOf[subschema #1]/items/gender_confidence` to the response with the `200` status
  - added the optional property `call_log/evaluations/anyOf[subschema #1]/items/language_confidence` to the response with the `200` status
  - …2 more
- **2026-08-08** `f48cef80963f` — 14 warning
  - removed the optional property `call_log/evaluations/anyOf[subschema #1]/items/agent_audio_interfering_speech` from the response with the `200` status
  - removed the optional property `call_log/evaluations/anyOf[subschema #1]/items/agent_audio_media_speech` from the response with the `200` status
  - removed the optional property `call_log/evaluations/anyOf[subschema #1]/items/agent_audio_noise` from the response with the `200` status
  - removed the optional property `call_log/evaluations/anyOf[subschema #1]/items/agent_audio_packet_loss` from the response with the `200` status
  - …10 more
- …earlier changes not shown

[Full history](https://skmtc.dev/getbluejay/apis/bluejay-api/changes/v1/retrieve-call-log/:call_id/get.md)

---

[API](https://skmtc.dev/getbluejay/apis/bluejay-api.md) · [All operations](https://skmtc.dev/getbluejay/apis/bluejay-api/llms.txt) · [OpenAPI document](https://skmtc-service-production.skmtc.workers.dev/v1/apis/getbluejay/bluejay-api/revisions/bf69ab7268a4/schema)