Data export service

You can leverage the Gatekeeper Data Export service to extract raw data and build the copy of the entire Gatekeeper fields into an in-house database. This can be achieved by leveraging a standard ETL tool, such as Azure Data Factory, or others equivalents.

Export data

Gatekeeper offers a mechanism to export baseline data for typical ETL process. Click here to get the data schema and the documentation of each attribute.

The entire schema is offered as a set of CSV files representing the following:

• Entity attributes
• Operation attributes
• Factor result attributes

Entity attributes

Following are the entity schema files. The entity schema files have the ent- prefix added to their name:

• ent-engagements
• ent-fraudsters
• ent-groups-persons
• ent-groups
• ent-media-segments
• ent-persons
• ent-processed-audio
• ent-sessions
• ent-voiceprints-enrollment-segments
• ent-voiceprints
• ent-watchlists-fraudsters
• ent-watchlists

Operation attributes

Following are the operation schema files. The operation files have the opr- prefix added to their name:

• opr-delete-personal-data

Factor result attributes

Following are the factor results files. The result files have the res- prefix added to their name:

• res-channel-playback-detections
• res-detect_known-fraudsters-voiceprint
• res-detect-known-fraudsters-convoprint-details
• res-detect-known-fraudsters-voiceprint-details
• res-enroll-convoprint
• res-enroll-voiceprint
• res-identify-convoprint-details
• res-identify-convoprint
• res-identify-voiceprint-details
• res-identify-voiceprint
• res-lock-convoprint
• res-lock-voiceprint
• res-opt-person-in
• res-opt-person-out
• res-process-audio-for-enroll
• res-process-audio
• res-process-text
• res-record-media
• res-sessions-factors
• res-sessions-involved-fraudsters
• res-sessions-involved-persons
• res-suspicious-session
• res-synthetic-speech-detection
• res-train-voiceprint
• res-unlock-convoprint
• res-unlock-voiceprint
• res-validate-text
• res-verify-convoprint
• res-verify-voiceprint

Nuance recommends using the schema CSV files as a starting point to build an SQL relational database where each CSV represents a separate table, and each table can be linked using foreign keys. For example, they can link the factors to gk_session_id and gk_engagement_id.

Schema headers

Each schema CSV file comes with headers, as mentioned below, so that customers can monitor possible changes to the schema:

• a versioned number
• a schema name (also known as <dataset-name>)
• a schema description
• a creation timestamp

The addition of new fields is not considered a breaking compatibility point.

General description of the schema columns

Generally the name of the schema file represent the name of a table in the destination SQL Database.

The following table describes fields mentioned in schema CSV files:

Download the data schema file

Download the latest data schema file from the following link:

• data schema v1.0.11

Export data output

Gatekeeper generates a zip output file that contains all the aggregated CSVs. The zip is generated every 5 minutes.

You can access that zip through an https URL pattern. For example, https://<Gatekeeper public API endpoint>/v1/data/gk_scope_id/<schema-name>/yyyy/MM/dd/HH/<schema-name>-yyyyMMddHHmm-yyyyMMddHHmm.zip.

The following tables lists more detailed examples:

HTTP response status codes

HTTP response status codes indicate if a specific HTTP request has been successfully completed. The x-ms-error-code header is available in case of error only. Following are common response codes and x-ms-error-code header value during client error:

For more details on HTTP status codes, see HTTP response status codes  .

General description of output columns

Each CSV output file has its own columns that are documented in the schema CSV. There are three special columns:

• $field_mask: List of fields used in the request - mainly important of an updated record, to determine which fields were updated among all other columns.
• $operation: An enum that describes which operation is performed:
  • 1 is for insert.
  • 2 is for update.
  • 3 is for delete.
• $timestamp: Original timestamp of the time when an event is published in Gatekeeper. Note that this timestamp is different from the timestamps visible in the web portal.
• $internal_reference_id: Contains information about the ID that helps in debugging what went wrong with the record.

Scope configuration parameters

The following are the configuration parameters to control export data:

• DataExport\EnableDataExport (Boolean): Default is false
• DataExport\ExportCsvHeaderPrefix (String): CSV header prefix for data export
• DataExport\ExportCustomData: Pop-up with two columns:
  • First column is a drop-down to select the schema name related to entities.
  • Second column is a text field to enter the list of comma-separated attributes. Each attribute in this field becomes a new column in the output CSV file.