Operations

Some operation parameters are required or should have a specific format, otherwise they’re unable to initiate the operation. If that happens, Gatekeeper reports an event of the OPERATION_ERROR type to the webhook. Errors that happen after the operation is launched, for example making reference to an unknown recordingId results with its corresponding *_COMPLETED_EVENT containing a non-OK status code.

Common parameters

A common parameter is configsetName: The configuration set name to use. It’s recommended you use the value stored in config.defaultConfigSetName or a configuration set name that makes sense for the operation being executed.

START_RECORDING

Starts the audio recording. This should result in a RECORDING_STARTED_EVENT or RECORDING_ERROR_EVENT.

There should only be one active recording in a call session at a time.

Parameters include:

• recordingId - Required. The recording ID that Nuance Media Manager uses to identify what audio stream to record. The specific value depends on the CTI implementation used.
• recordingSide - Required. The channel to record. Supported values are:
  • RECORDING_SIDE_1
  • RECORDING_SIDE_2
  • RECORDING_SIDE_ALL
• speakerChannel - Required, if recordingSide is RECORDING_SIDE_ALL: Determines which channel to use for biometric operations. Typically, the channel or side that has the caller’s audio. Value should be the side or channel number (an integer greater than 0).
• maxRecordingDuration - Specifies the maximum recording duration in seconds. Value must be an integer greater than 0. If not set, the recording continues until stopped by STOP_RECORDING.
• configsetName - See Common parameters above.

STOP_RECORDING

Stops the recording of the current call session. This should result in a RECORDING_ENDED_EVENT.

Parameters include:

• configsetName - See Common parameters.

PAUSE_RECORDING

Pauses the recording of the current call session. This should result in a RECORDING_PAUSED_EVENT.

Parameters include:

• configsetName - See Common parameters.

RESUME_RECORDING

Resumes a previously paused recording. This should result in a RECORDING_RESUMED_EVENT.

Parameters include:

• configsetName - See Common parameters.

GET_PERSON_STATUS

Gets the person status. This results in a PERSON_STATUS_EVENT.

Parameters include:

• externalPersonId - Required: The external person ID. This external ID should be unique for the current scope. Once found, the resulting internal person ID should be used in following operations.
• configsetName - See Common parameters.

GET_VOICEPRINT_STATUS

Gets the voiceprint enrollment status of the given (internal) person ID and voiceprint tag. This results in VOICEPRINT_STATUS_EVENT. If the voiceprint doesn’t already exist, an empty print is created, meaning result.VoiceprintStatus.status is set to VOICEPRINT_PROFILE_STATUS_NOT_TRAINED.

Parameters include:

• internalPersonId - Required. The internal person ID.
• voiceprintTag - The voiceprint tag to identify the voiceprint.
• configsetName - See Common parameters.

GET_CONVOPRINT_STATUS

Gets the convoprint enrollment status of the given (internal) person ID and convoprint tag. This results in CONVOPRINT_STATUS_EVENT.

If the conversation print doesn’t already exist and the createIfMissing parameter is set to true, an empty print is created, meaning result.voiceprintStatus.status is set to CONVOPRINT_PROFILE_STATUS_NOT_TRAINED and result.convoprintStatus.found is set to true. Otherwise, if an existing matching profile wasn’t found, result.convoprintStatus.found is set to false.

Parameters include:

• internalPersonId - Required. The internal person ID.
• convoprintTag - Required. The tag to identify the conversation print of the person to use. This is typically the language code.
• createIfMissing: Set to true to create a new empty convoprint profile if it doesn’t already exist.
• configsetName - See Common parameters.

GET_DEVICEPRINT_STATUS

Gets the deviceprint enrollment status of the given (internal) person ID and deviceprint tag. This results in DEVICEPRINT_STATUS_EVENT.

Parameters include:

• internalPersonId - Required. The internal person ID.
• deviceprintTag - Required. The tag to identify the conversation print of the person to use. This is typically the language code.
• configsetName - See Common parameters.

START_PROCESS_AUDIO_FOR_ENROLL

Starts to process the audio, identified by the mediaSegmentId, for the purpose of enrollment.

This results in one or more of the following events:

• AUDIO_PROCESSING_FOR_ENROLL_STARTED_EVENT
• AUDIO_PROCESSING_FOR_ENROLL_PROGRESS_EVENT
• AUDIO_PROCESSING_FOR_ENROLL_INTERMEDIATE_RESULT_EVENT
• AUDIO_PROCESSING_FOR_ENROLL_COMPLETED_EVENT
• AUDIO_PROCESSING_FOR_ENROLL_STOPPED_EVENT

Parameters include:

• mediaSegmentId - Required. The audio segment ID. It’s a value from result.recording.mediaId.
• configsetName - See Common parameters.

STOP_PROCESS_AUDIO_FOR_ENROLL

Stops (or interrupts) the audio processing for enrollment prior to its completion. Should result in either AUDIO_PROCESSING_FOR_ENROLL_STOPPED_EVENT or AUDIO_PROCESSING_FOR_ENROLL_COMPLETED_EVENT if the audio processing was already completed.

Parameters include:

• processingId - Required. This should be the value from result.audioProcessingForEnroll.processingId.
• configsetName - See Common parameters.

ENROLL_VOICEPRINT

Enrolls the processed audio into a voiceprint profile ID. This results with a VOICE_ENROLL_COMPLETED_EVENT.

Parameters include:

• processedAudioId - Required. This audio ID that’s the result of audio processing for enrollment. This should be the value from result.audioProcessingForEnroll.processedAudioId.
• voiceprintId - Required. The voiceprint profile ID. This should be the value from result.voiceprintStatus.voiceprintProfileId.
• configsetName - See Common parameters.

START_DETECT_AUDIO_SPOOFING

Start audio spoofing detection, which currently includes playback and synthetic speech detection.

This results in one or more of the following events:

• DETECT_AUDIO_SPOOFING_STARTED_EVENT
• DETECT_AUDIO_SPOOFING_INTERMEDIATE_EVENT
• DETECT_AUDIO_SPOOFING_PROGRESS_EVENT
• DETECT_AUDIO_SPOOFING_STOPPED_EVENT
• DETECT_AUDIO_SPOOFING_COMPLETED_EVENT

There is currently no operation that stops audio spoofing detection.

Parameters include:

• audioProcessingId - This or mediaSegmentId is required. The processed audio ID. Can be result.audioProcessingForEnroll.processedAudioId or result.verify.processedAudioId.

  Note:

  This doesn’t necessarily require the operation building that audio ID needs to be completed.

• mediaSegmentId - This or audioProcessingId is required. The recording media segment ID, from result.recording.mediaId.

• configsetName - See Common parameters.

START_DETECT_FRAUD_VOICEPRINT

Starts known fraudster detection (Detect Known Fraudsters, or DKF) by comparing the current audio with the voiceprints in the given watchlist ID that match the given tag.

This results in one or more of the following events:

• DETECT_KNOWN_FRAUDSTERS_STARTED_EVENT
• DETECT_KNOWN_FRAUDSTERS_INTERMEDIATE_EVENT
• DETECT_KNOWN_FRAUDSTERS_PROGRESS_EVENT
• DETECT_KNOWN_FRAUDSTERS_STOPPED_EVENT
• DETECT_KNOWN_FRAUDSTERS_COMPLETED_EVENT

Parameters include:

• processedAudioId - This or mediaSegmentId is required. The processed audio ID. Can be result.audioProcessingForEnroll.processedAudioId or result.verify.processedAudioId.

  Note:

  This doesn’t necessarily require the operation building that audio ID needs to be completed.

• mediaSegmentId - This or processedAudioId is required. The recording media segment ID, from result.recording.mediaId.

• watchlistId - Required. The fraudster watchlist ID to use.

• watchlistTag - Required. The voiceprint tag in that watchlist ID to use.

• configsetName - See Common parameters.

STOP_DETECT_FRAUD_VOICEPRINT

Stops (or interrupts) the Known Fraudster Detection operation.

Results with DETECT_KNOWN_FRAUDSTERS_STOPPED_EVENT. If the operation has already completed, this results with DETECT_KNOWN_FRAUDSTERS_COMPLETED_EVENT.

Parameters include:

• processingId - Required. The value of result.detectKnownFraudsters.processingId.
• configsetName - See Common parameters.

START_TRAIN_VOICEPRINT

Starts to train the enrolled voiceprint. The voiceprint must be fully trained to be usable for verification.

Results in one or more of those following events:

• TRAIN_STARTED_EVENT
• TRAIN_STOPPED_EVENT
• TRAIN_COMPLETED_EVENT

There is currently no operation that you can use to stop voiceprint training.

Parameters include:

• voiceprintId - Required. The voiceprint profile ID, which should be from result.voiceprintStatus.voiceprintProfileId.
• configsetName - See Common parameters.

START_VERIFY_VOICEPRINT

Starts voiceprint verification.

Results in one or more of these events:

• VERIFY_STARTED_EVENT
• VERIFY_STOPPED_EVENT
• VERIFY_PROGRESS_EVENT
• VERIFY_INTERMEDIATE_RESULT_EVENT
• VERIFY_COMPLETED_EVENT

Parameters include:

• mediaSegmentId - Required. The Audio Segment ID, from result.recording.mediaId.
• voiceprintId - Required. The voiceprint profile ID, which should be from result.voiceprintStatus.voiceprintProfileId.
• configsetName - See Common parameters.

STOP_VERIFY_VOICEPRINT

Interrupts the voiceprint verification. Results with VERIFY_STOPPED_EVENT. If the operation has already completed, this results with VERIFY_COMPLETED_EVENT.

Parameters include:

• voiceprintProcessingId - Required: The value of result.verify.processingId.
• configsetName - See Common parameters.

START_IDENTIFY_VOICEPRINT

Starts voiceprint identification.

This results in one or more of the following events:

• IDENTIFY_STARTED_EVENT
• IDENTIFY_STOPPED_EVENT
• IDENTIFY_PROGRESS_EVENT
• IDENTIFY_INTERMEDIATE_RESULT_EVENT
• IDENTIFY_COMPLETED_EVENT

Parameters include:

• mediaSegmentId - Required. The audio segment ID, from result.recording.mediaId.
• internalGroupId - This or voiceprintIds is required. The group ID to use, which should be from result.groupInformation.internalGroupId.
• voiceprintTag - Required. if internalGroupId is used: The voiceprint tag in that group ID to use.
• voiceprintIds - This or internalGroupId is required. A comma-delimited list of voiceprint profile ID, from result.voiceprintStatus.voiceprintProfileId fetched from each group member. Use this if the SUBMIT_DYNAMIC_GROUP action is used.
• configsetName - See Common parameters.

STOP_IDENTIFY_VOICEPRINT

Stops (or interrupts) a voiceprint identification. This results with IDENTIFY_STOPPED_EVENT. If the operation has already completed, this results with IDENTIFY_COMPLETED_EVENT.

Parameters include:

• identifyProcessingId - Required. The value of result.identify.processingId.
• configsetName - See Common parameters.

SAVE_AUTHENTICITY

Saves in the session information that the agent marked the session as authentic.

Parameters include:

• configsetName - See Common parameters.

SAVE_CONSENT

Saves the person consent information in the database. Typically called after getting the PERSON_CONSENT_EVENT from the agent’s actions.

Parameters include:

• consentProvided - Set to true if the person consented (or opted in), false otherwise. It should be the value of result.consent.provided.

• internalPersonId - Required. The internal person ID. It should be the value of result.consent.internalPersonId.

• optOutReason - If opted out, that’s result.consent.provided is set to false, the reason given to the agent. It should be the value of result.consent.optOutReason.

  Note:

  The optOutReason parameter is ignored if offerOptInDuration isn’t set.

• offerOptInDuration - Number of days as an integer, after which the agent should offer to opt the person in again. Should be the value of result.consent.offerOptIn.

  Note:

  The offerOptInDuration parameter is ignored if optOutReason isn’t set.

• configsetName - See Common parameters.

SAVE_SESSION_SUSPICIOUS

Saves the suspicious information for this session. Typically called after getting the SESSION_SUSPICIOUS_EVENT.

Parameters include:

• isSuspicious - Set to true if the call is marked as suspicious, false otherwise. It should be the value of result.suspicious.
• suspiciousReason - The reason entered by the agent. It should be the value of result.suspiciousReason.
• configsetName - See Common parameters.

GET_SESSION_DECISION

Gets the current session decision. This should result in the event SESSION_DECISION_EVENT.

The session decision changes over time as biometric operations are executed.

CONVEY_SESSION_DECISION

Communicates the final session decision for the call to the agent.

Parameters include:

• sessionDecision - Required. The session decision. Should be the value of result.sessionDecision.decision. See SESSION_DECISION_EVENT for possible values.
• decisionReason - The reason for the decision. Should be the value of result.sessionDecision.reason. See SESSION_DECISION_EVENT for possible values.
• sessionDecisionRisk - The session’s risk score. Should be the value of result.sessionDecision.risk.
• sessionDecisionReliability - The session’s reliability score. Should be the value of result.sessionDecision.reliability.
• decisionExcludedFactorReasonName - Should be the value of result.sessionDecision.excludedFactor.reasonName, if any.
• decisionExcludedFactorReasonType - Should be the value of result.sessionDecision.excludedFactor.reasonType, if any.

REPORT_ERROR

Reports to the Web Agent Console that an error prevents the flow to execute properly. This displays an error screen to the agent, with the desiredAction and errorMessage parameters that can be seen in the Error Details part of the screen.

Parameters include:

• desiredAction - Required. The action ID that failed. Set to a valid action ID or UNDEFINED if not applicable.
• errorMessage - The error message.

DISPLAY_MESSAGE

Display an informational message in the Web Agent Console, or customize the message that’s displayed for a given session decision.

The call status area is used by the Web Agent Console to convey the current state of the call flow. If titleText or messageText is set to a non-empty value, the call status area is updated to display that text, using the style given in messageType. The message remains until another call flow event or DISPLAY_MESSAGE operation updates the call status area.

If titleText or messageText is set to a non-empty value, the parameters disableActions, disableRestart, and agentActionsToBlock can also be used to prevent some agent actions in the Web Agent Console until the end of the call session.

The text to display in the call status area for a given session decision can be changed using the decision parameters. This should be set before the session decision occurs in the call, and has effect only for the current call session.

There is an additional reserved text area that supports longer text and more formatting options. This can be changed using the customText parameters. The reserved text area isn’t impacted by the call flow event messages displayed in the call status area.

Parameters include:

• messageType - Required if titleText or messageText is set. The style of the message. It must be set to one of the following: info, warning, or error.
• titleLabel - The localization label for the call status area’s title. Specify a path defined in the JSON object of EnglishCustomLocalization. For example, CONSTANT.MESSAGES.TITLE1.
• titleText - The non-localized title’s text to show in the call status area. This text is used if the titleLabel parameter isn’t set or isn’t found in EnglishCustomLocalization. Setting this to a string that contains a single space character clears the title.
• messageLabel - The localization label for the call status area’s text. Specify a path defined in the JSON object of EnglishCustomLocalization. For example, CONSTANT.MESSAGES.TEXT1.
• messageText - The non-localized message’s text to show in the call status area. This text is used if the messageLabel parameter isn’t set or isn’t found in EnglishCustomLocalization. Setting this to a string that contains a single space character clears the message.
• disableActions - If true, disables the elements in the Web Agent Console that let the agent perform actions until the end of the current call, except for restart audio, which is controlled by the disableRestart parameter. This parameter is ignored unless titleText or messageText is set to a non-empty value.
• disableRestart - If true, disables the restart audio action in the Web Agent Console. This parameter is ignored unless titleText or messageText is set to a non-empty value.
• agentActionsToBlock - A comma-delimited list of actions to block in the Web Agent Console. If set, the disableActions and disableRestart parameters are ignored. This parameter is ignored unless titleText or messageText is set to a non-empty value. The following is the list of action values:
  • RESTART - Blocks the restart audio action.
  • OPT_IN - Blocks the opt in (or enroll) action.
• customText.titleArea.text: The title to display in the reserved text area. Setting this to a string that contains a single space character clears the title.
• customText.titleArea.label: The label for the title of the reserved text area, than can be used for localization.
• customText.titleArea.bold: If set to true, displays the title of the reserved text area as bold.
• customText.titleArea.italic: If set to true, displays the title of the reserved text area in italic.
• customText.titleArea.color: The color of the title in the reserved text area. It must be set to one of the following values:
  • PRIMARY: Primary, mapped to blue.
  • SECONDARY: Secondary. mapped to gray.
  • SUCCESS: Success, mapped to green.
  • DANGER: Danger, mapped to red.
  • WARNING: Warning, mapped to yellow.
  • INFO: Info, mapped to light blue.
  • LIGHT: Light, mapped to white.
  • DARK: Dark, mapped to black.
• customText.textArea.text: The text to display in the reserved text area. Setting this to a string that contains a single space character clears the text.
• customText.textArea.label: The label for the text of the reserved text area, than can be used for localization.
• customText.textArea.bold: If set to true, displays the text of the reserved text area as bold.
• customText.textArea.italic: If set to true, displays the text of the reserved text area in italic.
• customText.textArea.color: The color of the text in the reserved text area. It must be set to one of the following values:
  • PRIMARY: Primary, mapped to blue.
  • SECONDARY: Secondary. mapped to gray.
  • SUCCESS: Success, mapped to green.
  • DANGER: Danger, mapped to red.
  • WARNING: Warning, mapped to yellow.
  • INFO: Info, mapped to light blue.
  • LIGHT: Light, mapped to white.
  • DARK: Dark, mapped to black.
• decision.DECISIONTYPE.text: The text that should be used for DECISIONTYPE. The DECISIONTYPE must be replaced with the decision type, as described below.
• decision.DECISIONTYPE.label: The label to use for DECISIONTYPE than can be used for localization. The DECISIONTYPE must be replaced with the decision type, as described below.

For decision.DECISIONTYPE.label and decision.DECISIONTYPE.text, the DECISIONTYPE should be one of the following:

• AUTHENTIC
• NO_MATCH
• FRAUD
• UNCERTAIN
• FAILED
• NO_RISK_DETECTED

See SESSION_DECISION_EVENT for a description of those decision types.

For example, if you want to set the display text for the “Fraud” decision, you should set the parameter decision.FRAUD.text.

BLOCK_AGENT_ACTIONS

Indicates to the Web Agent Console that some actions should be unavailable to the agent until the call ends or until they are enabled back again with ENABLE_AGENT_ACTIONS. Blocked actions can’t be enabled back again before the end of the call. This doesn’t prevent other Call Session Manager (CSM) clients, the built-in flow, or a webhook from performing those actions.

Parameters include:

• agentActionsToBlock - Required. A comma-delimited list of actions to block in the Web Agent Console. The list of action values are the following:
  • RESTART - Blocks the restart audio action.
  • SUSPICIUS - Blocks the mark as suspicious action.
  • SUSPICIOUS - Blocks the mark as suspicious action.
  • UPDATE_PERSON_INFORMATION - Blocks updating the person information, such as the first and last name.
  • OPT_IN - Blocks the opt in (or enroll) action.
  • OPT_OUT - Blocks the opt out action.
  • AUTHENTICITY - Blocks the mark as authentic action.
  • PERSON_MANUALLY_IDENTIFIED - Blocks the manual identification of the person in a group.

ENABLE_AGENT_ACTIONS

Indicates to the Web Agent Console that some actions should be made available again.

Parameters include:

• agentActionsToEnable - Required. A comma-delimited list of actions to enable in the Web Agent Console. The list of action values are the following:
  • RESTART - Blocks the restart audio action.
  • SUSPICIUS - Blocks the mark as suspicious action.
  • SUSPICIOUS - Blocks the mark as suspicious action.
  • UPDATE_PERSON_INFORMATION - Blocks updating the person information, for example the first and last name.
  • OPT_IN - Blocks the opt in (or enroll) action.
  • OPT_OUT - Block the opt out action.
  • AUTHENTICITY - Block the mark as authentic action.
  • PERSON_MANUALLY_IDENTIFIED - Block the manual identification of the person in a group.

UPDATE_PERSON_STATUS

Displays the given person information in the Web Agent Console and optionally save it to the Gatekeeper database. Only information specified in the parameters is updated.

Parameters include:

• internalPersonId - Required: The internal person ID.
• wacDisplayOnly - Set to true to avoid saving the updated information in Gatekeeper. By default, this parameter is set to false and the information is saved in Gatekeeper.
• firstName - The first name. This must contain only letters (Unicode General Category “Letter”), spaces (U+0020), single quotes (U+0027), or dashes (U+002D), with a maximum length of 32 characters.
• lastName - The last name. This must contain only letters (Unicode General Category “Letter”), spaces (U+0020), single quotes (U+0027), or dashes (U+002D), with a maximum length of 32 characters.
• genderIdentity - The gender. Supported values are:
  • GENDER_MALE
  • GENDER_FEMALE
  • GENDER_UNKNOWN
  • GENDER_NON_BINARY
• preferredLanguage - The preferred spoken language.
• displayName - The free-form text to display the full name. This isn’t saved in Gatekeeper.
• additionalInformation - The free-form text to display alongside the person. This isn’t saved in Gatekeeper.
• customData.* - Every parameter that starts with the prefix customData. adds some custom application key and value to the person information in Gatekeeper. For example the parameter customData.k1 with value v1 sets the person data key k1 to the value v1.

POST_CALL_OPERATION

Performs the given postCallAction and its parameters after the call ends.

Since the operation starts and completes after the call session ID becomes invalid, the events related to that operation aren’t sent.

Parameters include:

• postCallAction - Required. The post-call operation to perform. Currently, only START_DETECT_FRAUD_VOICEPRINT is supported.
• Additional parameters applicable to postCallAction: See above.

SCHEDULE_EVENT

Schedules a future event.

After delay_seconds, a SCHEDULED_EVENT is sent with the given id.

All scheduled events that are still active when the call ends are canceled.

At the most, five concurrent scheduled events may be active per call at a time.

Parameters include:

• delaySeconds - Required. How long to wait before sending the event, in seconds. The value must be an integer between 0 and 86400 (one day).
• id - A value used to identify the SCHEDULED_EVENT. Should fit in Java’s int type (signed 32-bit integer).

SET_CTI_CUSTOM_DATA

Sets the given custom key and value as attached custom data in the CTI.

Parameters include:

• setKvpCallIdentifier - Required. The call identifier that’s used to write the custom data. This should be visible somewhere in cti.kvps.*, but the specific value depends on the CTI implementation used.
• key - Required. The custom data key to set.
• value - The custom data value to set. If not set, the value for the given key is cleared.

SET_ENGAGEMENT_CUSTOM_DATA

Saves one or more custom data key and value to the current engagement.

Parameters include:

• engagementId - Required. The internal engagement ID from state.engagementId.
• configsetName - See Common parameters.
• key - Required. The custom data key to set.
• value - Required. The custom data value to set.

SET_SESSION_CUSTOM_DATA

Saves one or more custom data key and value to the current session.

Parameters include:

• sessionId - Required: The internal session ID, not the call session ID, from state.sessionId.
• configsetName - See Common parameters.
• key - Required. The custom data key to set.
• value - Required. The custom data value to set.

SET_VOICEPRINT_CUSTOM_DATA

Saves one or more custom data key and value to a voiceprint profile.

Parameters include:

• voiceprintId - Required. The voiceprint profile ID from result.voiceprintStatus.voiceprintProfileId.
• configsetName - See Common parameters.
• key - Required. The custom data key to set.
• value - Required. The custom data value to set.

SET_CONVOPRINT_CUSTOM_DATA

Saves one or more custom data key and value to a conversation print profile.

Parameters include:

• convoprintId - Required. The conversation print profile ID from result.convoprintStatus.convoprintProfileId.
• configsetName - See Common parameters.
• key - Required. The custom data key to set.
• value - Required. The custom data value to set.

SET_PERSON_CUSTOM_DATA

Saves one or more custom data key and value to a person ID.

Parameters include:

• internalPersonId - Required. The internal person ID from result.personStatus.internalPersonId.
• configsetName - See Common parameters.
• key - Required. The custom data key to set.
• value - Required. The custom data value to set.

SET_GROUP_CUSTOM_DATA

Saves one or more custom data key and value to a group ID.

Parameters include:

• internalGroupId - Required. The internal group ID.
• configsetName - See Common parameters.
• key - Required. The custom data key to set.
• value - Required. The custom data value to set.

SET_MEDIA_SEGMENT_CUSTOM_DATA

Saves one or more custom data key and value to a media segment ID.

Parameters include:

• mediaSegmentId - Required. The media segment ID from result.recording.mediaId.
• configsetName - See Common parameters.
• key - Required. The custom data key to set.
• value - Required. The custom data value to set.

SET_PROCESSED_AUDIO_CUSTOM_DATA

Saves one or more custom data key and value to a processed audio ID.

Parameters include:

• processedAudioId - Required. The processed audio ID from result.audioProcessingForEnroll.processedAudioId or result.verify.processedAudioId.
• configsetName - See Common parameters.
• key - Required. The custom data key to set.
• value - Required. The custom data value to set.

GET_GROUP_INFORMATION

Gets the group membership information and triggers a GROUP_INFORMATION_EVENT event. If an existing group isn’t found, an empty group is created.

Parameters include:

• externalGroupId - Required. The external group identifier.
• configsetName - See Common parameters.

SUBMIT_DYNAMIC_GROUP

Submits the information about a dynamic group that’s used for identification. Use this instead of GET_GROUP_INFORMATION, if you don’t want to use or create a group entity in the Gatekeeper database.

This event is followed by the GET_PERSON_STATUS and UPDATE_PERSON_STATUS actions for every external person ID.

Parameters include:

• externalGroupId - Required. The external group identifier, only used by the WAC for display.
• externalPersonId.N - Required. The external person IDs, where N is a decimal integer that starts with 0 and increases by 1 for every new member.
• tooBig - Set to true if the dynamic group is too big and the list of externalPersonId.N parameters have been truncated.
• configsetName - See Common parameters.

UPDATE_GROUP

Updates the list of members in a group.

Parameters include:

• internalGroupId - Required. The internal group ID from result.groupInformation.internalGroupId.
• internalPersonIds - Required. A comma-delimited (",") list of internal person IDs. This replaces the existing group membership with this list.

MANUAL_IDENTIFY_PERSON

Identifies the current speaker as the given person ID using some external means.

This is used for display only. Enrollment process is done using START_PROCESS_AUDIO_FOR_ENROLL and ENROLL_VOICEPRINT using a voiceprintId that corresponds to the identified speaker.

Parameters include:

• externalPersonId - The identified speaker’s person ID.

PREVIOUS_SESSION_RESULT

Conveys the results of the previous session in the current engagement to the agent.

See result.previousSession prefix for information about how to retrieve the previous session’s variables.

Parameters include:

• externalPersonId - The external person ID used for authentication.
• personManuallyIdentified - Set to true if the externalPersonId is manually identified by the agent, prior to START_VERIFY_VOICEPRINT or after IDENTIFY_COMPLETED_EVENT that returned a mismatch.
• enrollStatus - The value of result.previousSession.result.enroll.basicResult.status.
• enrollReason - The value of result.previousSession.result.enroll.basicResult.reason.
• trainStatus - The value of result.previousSession.result.train.basicResult.status.
• trainVoiceprintProfileStatus - The value of result.previousSession.result.train.basicResult.voiceprintProfileStatus.
• trainReason - The value of result.previousSession.result.train.basicResult.reason.
• sessionDecision - The value of result.previousSession.result.sessionDecision.decision.
• sessionDecisionReason - The value of result.previousSession.result.sessionDecision.reason.
• sessionDecisionRisk - The value of result.previousSession.result.sessionDecision.risk.
• sessionDecisionReliability - The value of result.previousSession.result.sessionDecision.reliability.

RESTART_SESSION

Creates a new call session and Gatekeeper session for the active call. The current call session gets RESTART_END_SESSION_EVENT, and the new call session gets RESTART_NEW_SESSION_EVENT. After these events are sent, only the new call session interacts with the Web Agent Console.

The new session is created within the same engagement as the current session, unless createNewEngagement is set to true. A new engagement should be created if the person identified as the speaker has changed, because of a new speaker, or the initial identification was incorrect.

Parameters include:

• createNewEngagement - Set to true if the new session uses a new engagement. Otherwise, the new session remains in the current engagement. Default is false.
• reAuthentication - Set to true if the new call session is for the same external person ID and should use the same Gatekeeper session. If true, the same engagement is used and the createNewEnagegement parameter is ignored. Default is false.

VALIDATE_CALLER_ID

Performs the caller ID validation. This results with the VALIDATE_CALLER_ID_RESULT_EVENT event.

Parameters include:

• configsetName - See Common parameters.
• referenceId: The reference ID that’s used with the Caller Identification Service’s ProcessPreAnswer. Typically, this should be the first engagement ID used in the current call.

START_TEXT_PROCESSING

Starts transcribing the audio into text for the purpose of creating a conversation print (“convoprint”).

This results in one or more of the following events:

• TEXT_PROCESSING_STARTED_EVENT
• TEXT_PROCESSING_PROGRESS_EVENT
• TEXT_PROCESSING_INTERMEDIATE_RESULT_EVENT
• TEXT_PROCESSING_COMPLETED_EVENT
• TEXT_PROCESSING_STOPPED_EVENT

Parameters include:

• mediaSegmentId - Required. The audio segment ID. It’s usually the value from result.recording.mediaId.
• language: Language and country (locale) code as xx-XX or xxx-XXX, such as en-US or eng-USA. If not specified, the method uses the actual_language processed audio attribute, or the configuration parameter. At least one of them must be valid. To know more about the list of language codes that we support, see Geographies  .
• internalPersonId: Owner of the audio or text. When set, the system handles the audio or text as personal information.
• configsetName - See Common parameters.

STOP_TEXT_PROCESSING

Stops (or interrupts) the transcription. It should result in either TEXT_PROCESSING_STOPPED_EVENT or TEXT_PROCESSING_COMPLETED_EVENT if the transcription is already completed just prior.

Parameters include:

• processingId - Required. This should be the value from result.textProcessing.processingId.
• configsetName - See Common parameters.

ENROLL_CONVOPRINT

Enrolls the transcribed audio into a convoprint ID. This results with CONVOPRINT_ENROLL_COMPLETED_EVENT.

Parameters include:

• convoprintId - Required. The conversation print profile ID, which should be from result.convoprintStatus.convoprintProfileId.
• textId - Required. The transcribed text ID, typically from result.textProcessing.textId.
• configsetName - See Common parameters.

START_DETECT_FRAUD_CONVOPRINT

Starts known fraudster detection (Detect Known Fraudsters, or DKF) by trying to match the current audio with the conversation prints in the given watchlist ID that match the given tag.

This results in one or more of those following events:

• DETECT_KNOWN_FRAUDSTERS_CONVOPRINT_STARTED_EVENT
• DETECT_KNOWN_FRAUDSTERS_CONVOPRINT_INTERMEDIATE_EVENT
• DETECT_KNOWN_FRAUDSTERS_CONVOPRINT_PROGRESS_EVENT
• DETECT_KNOWN_FRAUDSTERS_CONVOPRINT_STOPPED_EVENT
• DETECT_KNOWN_FRAUDSTERS_CONVOPRINT_COMPLETED_EVENT

Parameters include:

• processedAudioId - This or mediaSegmentId or textId is required. The processed audio ID.

  Note:

  This doesn’t necessarily require the operation building that audio ID needs to be completed.
• mediaSegmentId - This or processedAudioId or textId is required. The recording media segment ID from result.recording.mediaId.
• textId - This or processedAudioId or mediaSegmentId is required. The transcribed text ID.
• watchlistId - Required. The fraudsters watchlist ID to use.
• watchlistTag - Required. The conversation print tag that the watchlist ID uses. This is typically the language code.
• configsetName - See Common parameters.

STOP_DETECT_FRAUD_CONVOPRINT

Stops (or interrupts) the known fraudster detection. This should result in either DETECT_KNOWN_FRAUDSTERS_CONVOPRINT_STOPPED_EVENT or DETECT_KNOWN_FRAUDSTERS_CONVOPRINT_COMPLETED_EVENT if the operation has already completed just prior.

Parameters include:

• processingId - Required. The value of result.detectKnownFraudstersConvoprint.processingId.
• configsetName - See Common parameters.

START_TRAIN_CONVOPRINT

Starts to train the enrolled conversation print. The conversation print must be fully trained to be usable for conversation print verification.

This results in one or more of those following events:

• CONVOPRINT_TRAIN_STARTED_EVENT
• CONVOPRINT_TRAIN_STOPPED_EVENT
• CONVOPRINT_TRAIN_COMPLETED_EVENT

There is no operation that you can use to stop convoprint training.

Parameters include:

• convoprintId - Required. The conversation print profile ID, which should be from result.convoprintStatus.convoprintProfileId.
• configsetName - See Common parameters.

START_VERIFY_CONVOPRINT

Starts the conversation print verification.

This results in one or more of those following events:

• CONVOPRINT_VERIFY_STARTED_EVENT
• CONVOPRINT_VERIFY_STOPPED_EVENT
• CONVOPRINT_VERIFY_PROGRESS_EVENT
• CONVOPRINT_VERIFY_INTERMEDIATE_RESULT_EVENT
• CONVOPRINT_VERIFY_COMPLETED_EVENT

Parameters include:

• processedAudioId - This or mediaSegmentId or textId is required. The processed audio ID.

  Note:

  This doesn’t necessarily require the operation building that audio ID needs to be completed.
• mediaSegmentId - This or processedAudioId or textId is required. The recording media segment ID from result.recording.mediaId.
• textId - This or processedAudioId or mediaSegmentId is required. The transcribed text ID.
• convoprintId - Required. The conversation print profile ID, which should be from result.convoprintStatus.convoprintProfileId.
• language: Language and country (locale) code as xx-XX or xxx-XXX, such as en-US or eng-USA. If not specified, the method uses the actual_language processed audio attribute, or the configuration parameter. At least one of them must be valid. To know more about the list of language codes that we support, see Geographies  .
• configsetName - See Common parameters.

STOP_VERIFY_CONVOPRINT

Stops (or interrupts) the conversation print verification. This should result in either CONVOPRINT_VERIFY_STOPPED_EVENT or CONVOPRINT_VERIFY_COMPLETED_EVENT if the operation has already completed just prior.

Parameters include:

• convoprintProcessingId - Required. The value of result.convoprintVerify.processingId.
• configsetName - See Common parameters.

START_PROCESS_AUDIO_DEVICEPRINT

Starts processing the audio for the purpose of enrolling a deviceprint.

This results in one or more of the following events:

• AUDIO_PROCESSING_DEVICEPRINT_STARTED_EVENT
• AUDIO_PROCESSING_DEVICEPRINT_STOPPED_EVENT
• AUDIO_PROCESSING_DEVICEPRINT_PROGRESS_EVENT
• AUDIO_PROCESSING_DEVICEPRINT_INTERMEDIATE_RESULT_EVENT
• AUDIO_PROCESSING_DEVICEPRINT_COMPLETED_EVENT

Parameters include:

• mediaSegmentId - Required. The audio segment ID. It’s usually the value from result.recording.mediaId.
• configsetName - See Common parameters.

STOP_PROCESS_AUDIO_DEVICEPRINT

Stops or interrupts the audio processing for deviceprint enrollment prior to its completion. Should result in either AUDIO_PROCESSING_DEVICEPRINT_STOPPED_EVENT or AUDIO_PROCESSING_DEVICEPRINT_COMPLETED_EVENT if the audio processing was already completed.

Parameters include:

• processingId - Required. This should be the value from result.audioProcessingDeviceprint.processingId.
• configsetName - See Common parameters.

ENROLL_DEVICEPRINT

Enrolls the processed deviceprint audio into a deviceprint ID. This results with DEVICEPRINT_ENROLL_COMPLETED_EVENT.

Parameters include:

• deviceprintId - Required. The deviceprint profile ID, which should be from result.deviceprintStatus.deviceprintProfileId.
• processedAudioId - Required. The processed deviceprint audio ID, typically from result.audioProcessingDeviceprint.processedAudioId.
• configsetName - See Common parameters.

START_TRAIN_DEVICEPRINT

Starts to train the enrolled deviceprint. It must be fully trained to be usable for deviceprint verification.

This results in one or more of those following events:

• DEVICEPRINT_TRAIN_STARTED_EVENT
• DEVICEPRINT_TRAIN_STOPPED_EVENT
• DEVICEPRINT_TRAIN_COMPLETED_EVENT

There is no operation that you can use to stop deviceprint training.

Parameters include:

• deviceprintId - Required. The conversation print profile ID, which should be from result.convoprintStatus.convoprintProfileId.
• configsetName - See Common parameters.

START_VERIFY_DEVICEPRINT

Starts the deviceprint verification.

This results in one or more of those following events:

• DEVICEPRINT_VERIFY_STARTED_EVENT
• DEVICEPRINT_VERIFY_STOPPED_EVENT
• DEVICEPRINT_VERIFY_PROGRESS_EVENT
• DEVICEPRINT_VERIFY_INTERMEDIATE_RESULT_EVENT
• DEVICEPRINT_VERIFY_COMPLETED_EVENT

Parameters include:

• processedAudioId - This or mediaSegmentId is required. The processed audio ID. Note that this doesn’t necessarily require the operation building that audio ID needs to be completed.
• mediaSegmentId - This or processedAudioId is required. The recording media segment ID from result.recording.mediaId.
• deviceprintId - Required. The deviceprint profile ID, which should be from result.deviceprintStatus.deviceprintProfileId.
• configsetName - See Common parameters.

STOP_VERIFY_DEVICEPRINT

Interrupts the deviceprint verification. Results with DEVICEPRINT_VERIFY_STOPPED_EVENT. If the operation has already completed, this results with DEVICEPRINT_VERIFY_COMPLETED_EVENT.

Parameters include:

• deviceprintProcessingId - Required. The value of result.deviceprintVerify.processingId.
• configsetName - See Common parameters.

START_DETECT_FRAUD_DEVICEPRINT

Starts known fraudster detection (also known as Detect Known Fraudsters, or DKF) by trying to match the current audio with the deviceprints in the given watchlist ID that match the given tag.

This results in one or more of those following events:

• DETECT_KNOWN_FRAUDSTERS_DEVICEPRINT_STARTED_EVENT
• DETECT_KNOWN_FRAUDSTERS_DEVICEPRINT_INTERMEDIATE_EVENT
• DETECT_KNOWN_FRAUDSTERS_DEVICEPRINT_PROGRESS_EVENT
• DETECT_KNOWN_FRAUDSTERS_DEVICEPRINT_STOPPED_EVENT
• DETECT_KNOWN_FRAUDSTERS_DEVICEPRINT_COMPLETED_EVENT

Parameters include:

• processedAudioId - This or mediaSegmentId is required. The processed audio ID. Note that this doesn’t necessarily require the operation building that audio ID needs to be completed.
• mediaSegmentId - This or processedAudioId is required. The recording media segment ID from result.recording.mediaId.
• watchlistId - Required. The fraudsters watchlist ID to use.
• watchlistTag - The deviceprint tag in that watchlist ID to use. If not set, all deviceprints in the watchlist are used.
• configsetName - See Common parameters.

STOP_DETECT_FRAUD_DEVICEPRINT

Stops or interrupts the known fraudster detection. This should result in either DETECT_KNOWN_FRAUDSTERS_DEVICEPRINT_STOPPED_EVENT or DETECT_KNOWN_FRAUDSTERS_DEVICEPRINT_COMPLETED_EVENT if the operation has already completed just prior.

Parameters include:

• processingId - Required. The value of result.detectKnownFraudstersDeviceprint.processingId.
• configsetName - See Common parameters.

ASSESS_CALLER_ID_RISK

Perform a risk assessment of the caller ID again the caller ID watchlist. Results with ASSESS_CALLER_ID_EVENT.

Parameters include:

• callerId - Required. The caller ID, typically the value of cti.ani.
• watchlistId - The caller ID watchlist ID to use. If not set, the scope’s default caller ID watchlist is used.
• configsetName - See Common parameters.

ENABLE_AFTER_CALL_WORK_MODE

Indicate that the current call session should enter after-call work mode once the telephony call ends. Only the first of this operation in a call session has effect, as subsequent ENABLE_AFTER_CALL_WORK_MODE operations are ignored.

Parameters include:

• afterCallWorkModeDurationMs - Required. The maximum allowed duration of the after-call work mode, in milliseconds. The value must be greater than 0.

AFTER_CALL_WORK_MODE_DONE

Indicate that after-call work mode should now end. This is ignored unless it’s sent after AFTER_CALL_WORK_MODE_START_EVENT and before AFTER_CALL_WORK_MODE_END_EVENT.

Related topics

• Event types
• Session keys