Get traces | Sarvam API Docs

import requests
url = "https://apps.sarvam.ai/api/analytics/v1/acme-corp-7c2e9f/support-ws-4b1d8a/support-bot-3f9a1c2e-7b4d/traces/20261007%2F9f1c2a7e-4b3d-4e8a-a1f0-6c2d9b7e5a14"
headers = {"X-API-Key": "<your-api-key>"}
response = requests.get(url, headers=headers)
print(response.json())

{
  "interaction_id": "20261007/9f1c2a7e-4b3d-4e8a-a1f0-6c2d9b7e5a14",
  "lifecycle_events": [
    {
      "event_type": "ServerToolEvent.on_start",
      "display_name": "Conversation Started",
      "status": "success",
      "input": {
        "user_information": {
          "identifier": "+919876543210",
          "user_identifier_type": "phone"
        }
      },
      "output": {
        "agent_variables": {
          "customer_name": "Priya",
          "loan_status": "overdue"
        },
        "initial_bot_message": "Namaste Priya, this is Asha from Acme Finance."
      }
    },
    {
      "event_type": "ServerToolEvent.on_end",
      "display_name": "Conversation Ended",
      "status": "success",
      "input": {
        "agent_variables": {
          "customer_name": "Priya",
          "loan_status": "overdue",
          "payment_date": "2026-10-12"
        }
      },
      "output": {}
    }
  ],
  "turns": [
    {
      "turn_id": 1,
      "is_merged": false,
      "events": [
        {
          "event_type": "ServerASREvent",
          "display_name": "ASR",
          "status": "success",
          "input": {
            "audio_duration": 2.4,
            "language_tag": "hi-IN",
            "model_name": "saarika:v2.5",
            "prompt": ""
          },
          "output": {
            "language_prob": 0.97,
            "language_tag": "hi-IN",
            "text": "Haan, boliye."
          }
        },
        {
          "event_type": "ServerChatLLMEvent.assistant",
          "display_name": "Chat LLM",
          "status": "success",
          "input": {
            "text": "Haan, boliye."
          },
          "output": {
            "text": "Your EMI of ₹4,500 is overdue. When can you make the payment?"
          }
        },
        {
          "event_type": "ServerToolEvent.update_variables",
          "display_name": "Update Variables",
          "status": "success",
          "input": {
            "args": {
              "call_stage": "payment_reminder"
            }
          },
          "output": {
            "text": "Variables updated."
          }
        },
        {
          "event_type": "ServerTTSEvent",
          "display_name": "TTS",
          "status": "success",
          "input": {
            "assistant_text": "Aapki ₹4,500 ki EMI baaki hai. Aap payment kab kar payenge?",
            "language_tag": "hi-IN",
            "speaker_id": "anushka"
          },
          "output": {
            "language_tag": "hi-IN",
            "voice_settings": {
              "pace": 1,
              "pitch": 0,
              "speed": 1,
              "style": 0
            }
          }
        }
      ]
    }
  ]
}

Returns the event trace of an interaction: each turn with its speech recognition (ASR), LLM, translation, text-to-speech (TTS) and tool-call events, plus interaction-level events such as conversation start and end and post-call analysis. input and output vary by event_type. A trace can include personal data: the user’s contact details, agent variables and the transcript. Returns 404 if no trace exists.

curl "https://apps.sarvam.ai/api/analytics/v1/$ORG_ID/$WORKSPACE_ID/$APP_ID/traces/$INTERACTION_ID" \
  -H "X-API-Key: $SARVAM_API_KEY"

Path parameters

org_idstringRequired

Organization ID.

workspace_idstringRequired

Workspace ID.

app_idstringRequired

Agent ID. Copy it from the agent’s ··· menu in the dashboard (Copy agent ID).

interaction_idstringRequiredformat: "^[0-9]{8}/.+$"

From List interactions, in the form <YYYYMMDD>/<id>, e.g. 20261007/9f1c2a7e-4b3d-4e8a-a1f0-6c2d9b7e5a14. Send the slash as is or as %2F.

Response

The interaction's event trace.

interaction_idstring

The interaction this trace belongs to.

lifecycle_eventslist of objectsOptional

Interaction-level events: conversation start and end, their logs, post-call analysis, and errors.

turnslist of objectsOptional

The conversation's turns, in order, each with its events.

Errors

400

Bad Request Error

401

Unauthorized Error

403

Forbidden Error

404

Not Found Error

422

Unprocessable Entity Error

429

Too Many Requests Error

500

Internal Server Error

503

Service Unavailable Error

Path parameters

Headers

Response

Errors