Work with exported recordings in AWS S3 bucket

AI-generated summary

Genesys Cloud exports call recordings to AWS S3 buckets organized in a hierarchical folder structure by organization ID, year, month, day, hour, and conversation ID, with all files secured through AWS S3 server-side encryption at rest. Recordings are stored as .opus files for audio, .zip files for screen recordings, and .json files for associated metadata, with encryption configurable using Amazon S3-managed keys (SSE-S3), AWS-managed keys, or customer-provided keys through AWS Key Management Service (SSE-KMS). Each recording includes a corresponding JSON metadata file suffixed with "_metadata.json" containing searchable details such as start time, duration, media type, and various identifiers, with all timestamps recorded in UTC time. Bulk export can be performed automatically through Quality Management (QM) policies or manually via the recording bulk action API after configuring the AWS S3 recording bulk actions integration. The JSON schema defines properties for media type, time information, IDs, and file details to enable efficient searching and analysis of exported recordings. Metadata is available for policy-based, screen, and snippet recordings, capturing comprehensive recording details including media type (Call, Chat, Email, Message, Screen) and media subtype (Trunk, Station, Consult, Screen). Key technical fields within the metadata include unique identifiers for recordings, conversations, and organizations; temporal data such as start time, end time, and duration in milliseconds; and communication-specific attributes including ANI (Automatic Number Identification) and DNIS (Dialed Number Identification Service) in both normalized and displayable formats. Policy-based and screen recordings include arrays for multiple users, queue IDs, and wrap-up codes, while snippet recordings track individual user and queue identifiers. Additional metadata encompasses conversation direction (inbound/outbound), message platform types (SMS, Twitter, Facebook, WhatsApp, web messaging, Instagram), language identifiers, file paths, and file sizes. Screen recordings include specific screen information with screen ID, X and Y position coordinates, and resolution data. Required fields across all recording types include media type, provider, temporal markers, organization and conversation IDs, recording ID, file path, and file size. Systems with additional Recording Export Encryption enabled require manual decryption of file content after retrieval from the S3 bucket, though Genesys Cloud does not support object-level encryption via PGP in the FedRAMP region, where only server-side encryption is supported.

After the AWS S3 recording bulk actions integration is set up, you can export recordings that are in Genesys Cloud in bulk to your AWS S3 bucket. This export can be performed automatically via a QM policy, or explicitly by invoking the recording bulk action API.

This article details the content that is exported to your AWS S3 bucket.

AWS S3 bucket content

Recording files are exported to the AWS S3 bucket into folders with the following structure:

s3://{bucket}/{organizationId}/year={year}/{month={month}/day={day}/hour={hourOfDay}/conversation_id={conversationId}/

Placeholder	Description
{bucket}	The S3 bucket name.
{organizationId}	The organization ID.
{year}	The year in which the conversation began.
{month}	The month during which the conversation began (in digits).
{day}	The day during which the conversation began.
{hourOfDay}	The hour during which the conversation began.
{conversationId}	The conversation ID.

Note: The date and time during which the conversation began appears in UTC time.

The folder contains all the recording files that are retained during the conversation. Each recording file has one recording, and the name of the file is the recording ID.

Each recording file has a corresponding JSON metadata file. The JSON metadata file name is suffixed with “_metadata.json”.

The metadata can be used to search for the exported recording. For more information, see the Athena+Glue example (an example of a recording search service).

The metadata file is available in the JSON format for the following recordings.

JSON schema file (policy-based recording and screen recording)Click to expand

{
“$schema”: “http://json-schema.org/draft-04/schema#”,
“type”: “object”,
“properties”: {
“mediaType”: {
“description”: “Media Type (one of Call, Chat, Email, Message, Screen)”,
“type”: “string”
},
“mediaSubtype”: {
“description”: “The subtype of the recording (one of Trunk, Station, Consult, Screen)”,
“type”: “string”
},
“mediaSubject”: {
“description”: “The subject of the recording”,
“type”: “string”
},
“provider”: {
“description”: “Type of provider for the recording, for example edge”,
“type”: “string”
},
“userIds”: {
“description”: “List of users”,
“type”: “array”,
“items”: [
{
“type”: “string”
}
]
},
“startTime”: {
“description”: “Start time of the recordings”,
“type”: “string”
},
“endTime”: {
“description”: “End time of the recordings”,
“type”: “string”
},
“durationMs”: {
“description”: “Duration of the recording”,
“type”: “integer”
},
“initialDirection”: {
“description”: “Initial direction of the conversation (inbound/outbound)”,
“type”: “string”
},
“aniNormalized”: {
“description”: “ANI”,
“type”: “string”
},
“aniDisplayable”: {
“description”: “ANI in displayable form”,
“type”: “string”
},
“dnisNormalized”: {
“description”: “DNIS”,
“type”: “string”
},
“dnisDisplayable”: {
“description”: “DNIS in displayable form”,
“type”: “string”
},
“queueIds”: {
“description”: “List of queue IDs for the recording”,
“type”: “array”,
“items”: [
{
“type”: “string”
}
]
},
“wrapupCodes”: {
“description”: “Wrap-up codes for the conversation”,
“type”: “array”,
“items”: [
{
“type”: “string”
}
]
},
“organizationId”: {
“description”: “Unique ID for the conversation”,
“type”: “string”
},
“conversationId”: {
“description”: “Unique ID associated with the conversation”,
“type”: “string”
},
“conversationStartTime”: {
“description”: “Conversation’s start time”,
“type”: “string”
},
“conversationEndTime”: {
“description”: “Conversation’s end time”,
“type”: “string”
},
“recordingId”: {
“description”: “Unique ID for the recording”,
“type”: “string”
},
“filePath”: {
“description”: “Original path of the recording”,
“type”: “string”
},
“fileSize”: {
“description”: “Recording file size”,
“type”: “integer”
},
“messageType”: {
“description”: “Type of message platform from which the message originated, for example, SMS, twitter, line, Facebook, WhatsApp, web messaging, open, Instagram”,
“type”: “string”
},
“languageIds”: {
“description”: “Identifier on the language”,
“type”: “array”,
“items”: [
{
“type”: “string”
}
]
},
“screenInformation”: {
“description”: “Screen specific information, includes the screen ID, X and Y position, resolution information”,
“type”: “object”
}
},
“required”: [
“mediaType”,
“provider”,
“startTime”,
“endTime”,
“durationMs”,
“organizationId”,
“conversationId”,
“conversationStartTime”,
“conversationEndTime”,
“recordingId”,
“filePath”,
“fileSize”
]
}

JSON schema file (snippet recording)Click to expand

{
“$schema”: “http://json-schema.org/draft-04/schema#”,
“type”: “object”,
“properties”: {
“mediaType”: {
“description”: “Media Type (one of Call, Chat, Email, Message, Screen)”,
“type”: “string”
},
“mediaSubtype”: {
“description”: “The subtype of the recording (one of Trunk, Station, Consult, Screen)”,
“type”: “string”
},
“mediaSubject”: {
“description”: “The subject of the recording”,
“type”: “string”
},
“provider”: {
“description”: “Type of provider for the recording, for example edge”,
“type”: “string”
},
“userId”: {
“description”: “The identifier of the user from where the recording is originated”,
“type”: “string”
},
“startTime”: {
“description”: “Start time of the recordings”,
“type”: “string”
},
“endTime”: {
“description”: “End time of the recordings”,
“type”: “string”
},
“durationMs”: {
“description”: “Duration of the recording”,
“type”: “integer”
},
“initialDirection”: {
“description”: “Initial direction of the conversation (inbound/outbound)”,
“type”: “string”
},
“aniNormalized”: {
“description”: “ANI”,
“type”: “string”
},
“aniDisplayable”: {
“description”: “ANI in displayable form”,
“type”: “string”
},
“dnisNormalized”: {
“description”: “DNIS”,
“type”: “string”
},
“dnisDisplayable”: {
“description”: “DNIS in displayable form”,
“type”: “string”
},
“queueId”: {
“description”: “The identifier of the queue where the recording occurs”,
“type”: “string”
},
“organizationId”: {
“description”: “Unique ID for the conversation”,
“type”: “string”
},
“conversationId”: {
“description”: “Unique ID associated with the conversation”,
“type”: “string”
},
“conversationStartTime”: {
“description”: “Conversation’s start time”,
“type”: “string”
},
“recordingId”: {
“description”: “Unique ID for the recording”,
“type”: “string”
},
“filePath”: {
“description”: “Original path of the recording”,
“type”: “string”
},
“fileSize”: {
“description”: “Recording file size”,
“type”: “integer”
},
“languageId”: {
“description”: “Identifier on the language where the recording occurs”,
“type”: “string”
}
},
“required”: [
“mediaType”,
“provider”,
“startTime”,
“endTime”,
“durationMs”,
“organizationId”,
“conversationId”,
“conversationStartTime”,
“recordingId”,
“filePath”,
“fileSize”
]
}

For example, a call conversation with screen recording enabled may have the following folder content.

In the image below, the .opus file is the audio recording file, the .zip file contains the screen recording file, and the .json files are the JSON metadata associated with the respective media files.

Click the image to enlarge.

A digital conversation may have the following folder content.

In the image below, the .zip file contains the digital recording file, and the .json file is the corresponding JSON file.

Click the image to enlarge.

Encryption

Your S3 bucket is already configured with AWS S3 Server-Side Encryption (SSE). It may have been enabled with encryption keys that are Amazon S3 managed (SSE-S3), or enabled with the AWS-managed keys, or customer-provided keys from the AWS Key Management Service (SSE-KMS).

AWS S3 Server-Side Encryption (SSE) secures the recording files at rest in the S3 bucket. When the files are retrieved from the bucket, AWS automatically decrypts the file content.

If your system includes an additional enabled Recording Export Encryption, you must decrypt the file content yourself after you retrieve the files from the S3 bucket.

Note: Genesys does not support object level encryption on exported recordings via PGP in the FedRAMP region. Server-side encryption is supported.

[NEXT] Was this article helpful?

Get user feedback about articles.

Loading...

Work with exported recordings in AWS S3 bucket

AWS S3 bucket content

Encryption

[NEXT] Was this article helpful?