CTT File
The Contacter File (CTT) is a proprietary, structured JSON document that aggregates all transactional data, conversational analytics, and AI-generated metadata regarding a specific customer interaction. CTT files are natively initialized by the Transcriber engine and progressively augmented through downstream enrichment pipelines handled by the Contacter CLI or the Ainter orchestrator.
A standard CTT file encapsulates the following structural schemas:
File Naming Convention
CTT files are initialized using a standardized naming layout that encodes the four core primary keys of the customer interaction:
- Line In: The target called number or enterprise service boundary.
- Agent: The explicit identifier of the Customer Service Representative (CSR) who handled the interaction.
- Customer: The unique identification payload representing the calling customer.
- Date: The standardized date and time execution timestamp of the interaction.
The Transcriber module generates the baseline record (Version 0). Each subsequent enrichment layer increments the version tracking index. The file layout follows a strict tokenized format:
LineIn_Agent_Customer_Date_Version.ctt
UUID
A globally unique cryptographic identifier (UUID) assigned to map and isolate the call session seamlessly across external systems.
"uuid": "202a4e9e-052b-4365-9133-f873ee991359"
Tool
A fixed system string property defining the Contacter framework tracking version and schema signature.
"tool": "contacter"
Origin
An immutable metadata register capturing the physical properties of the source audio interaction, including the raw filename, channel/track count, baseline volume levels, and sample rates.
# Example of origin
"origin": {
"filein": "data/voice/16990_01_032_2018-07-30T153159.mp3",
"type": "audio",
"numchannelsorig": 1,
"samplerateorig": 16000,
"volumeorig": {
"0": 0.007
},
"durationorig": 1032.228
},
Process
Every workflow execution—whether an initial transcription or a downstream AI augmentation—is registered as an audit process. CTT files maintain a chronological array documenting the explicit configurations, runtime hyper-parameters, and execution results of each process block.
It also defines the specific schema components that were mutated during execution. Every modified component instantiates a new delta version containing only the updated variable states. Both the process header and the mutated component layers are cryptographically signed to ensure data integrity.
# Example of process
"process": {
"uuid": "a3575785-49ab-4bec-bc67-0cc7f3a2d7a8",
"operation": "ainter0.1",
"command": "src/contacter/ainter.py data/transcription/16990_01_032_2018-07-30T15:31:59_0.ctt data/identify
identify:stable -v 4",
"filein": "data/transcription/16990_01_032_2018-07-30T15:31:59_0.ctt",
"fileout": "data/identify/16990_01_032_2018-07-30T15:31:59_1.ctt",
"llm_metadata": {
"additional_kwargs": {
"refusal": null
},
"response_metadata": {
"token_usage": {
"completion_tokens": 6822,
"prompt_tokens": 7390,
"total_tokens": 14212,
"completion_tokens_details": {
"accepted_prediction_tokens": null,
"audio_tokens": 0,
"reasoning_tokens": 5658,
"rejected_prediction_tokens": null,
"image_tokens": 0
},
"prompt_tokens_details": {
"audio_tokens": 0,
"cached_tokens": 0,
"cache_write_tokens": 0,
"video_tokens": 0
},
"cost": 0,
"is_byok": false,
"cost_details": {
"upstream_inference_cost": 0,
"upstream_inference_prompt_cost": 0,
"upstream_inference_completions_cost": 0
}
},
"model_provider": "openai",
"model_name": "nex-agi/nex-n2-pro:free",
"system_fingerprint": null,
"id": "gen-1781806390-bCdjP59nsNJQPZCSdHiR",
"finish_reason": "stop",
"logprobs": null
},
"type": "ai",
"name": null,
"id": "lc_run--019edbef-da30-7422-908e-2889bbab6507-0",
"tool_calls": [],
"invalid_tool_calls": [],
"usage_metadata": {
"input_tokens": 7390,
"output_tokens": 6822,
"total_tokens": 14212,
"input_token_details": {
"audio": 0,
"cache_read": 0
},
"output_token_details": {
"audio": 0,
"reasoning": 5658
}
}
},
"processdate": "2026-06-18 18:14:41.575023+00:00",
"created": [
"speaker[1]"
],
"signature": "eqvACX68DzVl3LzU0Cnppf+Q4r7r1tVI...S7Dhrn1tkOGjHQV36ZBuUdXi70bRB+H+7TzKmQ=="
}
Project
A highly flexible, custom tracking layer used to segment data logs by enterprise campaigns, logical interaction groupings, distinct Contact Center site IDs, or any other tenant-specific categorization.
# Example of Project Component
"project": "myproject"
Main
The core operational ledger aggregating the fundamental routing vectors of the session, including the established Line, assigned Agent, Customer ID, and the validated date/time token.
# Example of Call Component
"main": {
"line": "16990",
"agent": "01",
"customer": "032",
"date": "2025-07-30T15:31:59"
}
Call
A global analytics matrix aggregating session-wide data points compiled across the overall interaction. This section includes details regarding independent channels, total conversational turns, absolute volume levels, cross-channel overlaps, language probability scores, and unified system processing metadata.
# Example of Call Component
"call": {
"channel": {
"innumber": 1
},
"turns": 149,
"volumeorig": {
"0": 0.007
},
"gaincorr": {
"0": 18.915
},
"volumecorr": {
"0": 0.118
},
"durationtranscription": 729.419,
"language": {
"en": {
"turns": 126,
"duration": 654.211,
"langprob": 0.967
},
"es": {
"turns": 2,
"duration": 2.514,
"langprob": 0.671
}
},
"mainlanguage": "en",
"nospeechprob": 0.0,
"logprob": -0.421,
"comratio": 1.081,
"trust": 65.64,
"wordnumber": 3384,
"overlap": {
"0": [
{
"start": 78.28,
"end": 78.854,
"duration": 0.574,
"speaker1": "SPEAKER_0001",
"speaker2": "SPEAKER_0005"
},
{
"start": 87.663,
"end": 88.287,
"duration": 0.624,
"speaker1": "SPEAKER_0001",
"speaker2": "SPEAKER_0005"
},
{
"start": 88.287,
"end": 88.658,
"duration": 0.371,
"speaker1": "SPEAKER_0001",
"speaker2": "SPEAKER_0004"
},
]
},
"tags": [
"mainspeaker=SPEAKER_0002",
"mainlanguage=en"
]
}
Speaker
An indexed repository tracking granular profiling metrics and acoustic data localized per speaker channel. It records total turns, spoken duration, voice amplitude parameters, overlap indicators, language metrics, heuristic behavioral tags, and associated process logs.
# Example of Speaker Component
"speaker": {
"SPEAKER_0001": {
"turns": 13,
"duration": 62.902,
"volumeorig": 0.006,
"volumecorr": 0.121,
"language": {
"en": {
"turns": 13,
"duration": 62.902,
"langprob": 0.992
}
},
"nospeechprob": 0.0,
"logprob": -0.289,
"comratio": 1.091,
"wordnumber": 134,
"overlap": {},
"trust": 74.9,
"tags": [
"mainlanguage=en",
"mainlanguagevalue=100",
"ANNOUNCE"
]
},
"SPEAKER_0005": {
"turns": 7,
"duration": 6.244,
"volumeorig": 0.009,
"volumecorr": 0.167,
"language": {
"en": {
"turns": 5,
"duration": 3.729,
"langprob": 0.56
},
"es": {
"turns": 2,
"duration": 2.514,
"langprob": 0.671
}
},
"nospeechprob": 0.0,
"logprob": -0.761,
"comratio": 0.539,
"wordnumber": 13,
"overlap": {},
"trust": 46.72,
"tags": [
"mainlanguage=en",
"mainlanguagevalue=60"
]
}
}
Dialog
A sequential timeline capturing every distinct dialogue entry (phrase segment) spoken during the interaction. This component maintains the exact start/end timestamps, individual turn indexes, phrase-level volume metrics, overlapping markers, transcription confidence flags, contextual dialogue tags, and the final raw text transcription payload.
# Example of Dialog Component
"dialog": [
{
"uuid": "c0a4be46-c41e-4e1e-873a-f83ab0e9b926",
"channel": 0,
"start": 2.174,
"end": 3.49,
"volumeorig": 0.009,
"volumecorr": 0.167,
"gaintranscription": 1.843,
"volumetranscription": 0.306,
"duration": 2.316,
"language": "en",
"langprob": 0.981,
"nospeechprob": 0.0,
"logprob": -0.227,
"comratio": 0.68,
"trust": 79.67,
"distortion": 0,
"speaker": "SPEAKER_0001",
"wordnumber": 3,
"overlay": [],
"initial_prompt": "",
"text": "Welcome to our Customer Center",
"tags": []
},
{
"uuid": "85dc148c-eeea-4bc5-b893-6715f3b7a1b3",
"channel": 0,
"start": 6.005,
"end": 15.944,
"volumeorig": 0.005,
"volumecorr": 0.097,
"gaintranscription": 2.509,
"volumetranscription": 0.238,
"duration": 9.939,
"language": "en",
"langprob": 0.999,
"nospeechprob": 0.0,
"logprob": -0.279,
"comratio": 1.143,
"trust": 75.67,
"distortion": 0,
"speaker": "SPEAKER_0001",
"wordnumber": 20,
"overlay": [],
"initial_prompt": "",
"text": "For memberships or commercial information, dial 1. If you are already a customer, dial 2. For technical support and scheduling, dial 5.",
"tags": []
},
{
"uuid": "ad509dd3-ffd6-45c4-bbfb-0da1efeeca5b",
"channel": 0,
"start": 17.024,
"end": 18.29,
"volumeorig": 0.006,
"volumecorr": 0.117,
"gaintranscription": 2.173,
"volumetranscription": 0.248,
"duration": 1.266,
"language": "pt",
"langprob": 1,
"nospeechprob": 0.0,
"logprob": -0.902,
"comratio": 0.733,
"trust": 40.58,
"distortion": 0,
"speaker": "SPEAKER_0001",
"initial_prompt": "",
"wordnumber": 4,
"overlay": [],
"forcelanguage": "en",
"text": "Jump with the spirit",
"tags": [
"UNCERTAIN_AUDIO"
]
}
]