Real-Time Speech Analytics Condition settings

Real-Time Speech Analytics Product that analyzes ongoing changes in customer behavior through spoken interactions with the customer, and drives effective organizational changes needed to address challenging market conditions. offers different options to monitor spoken terms, and define the conditions to be met in the Recorder Analytics Rule.

Real-Time Speech Analytics (RTSA Feature that detects keywords (terms and phrases) spoken in contact center interactions in near real time.) processing is segment-based. If interactions span multiple segments, the analytics process restarts for every segment, without any knowledge of prior segments. As a result, you may see issues with RTSA conditions, including but not limited to the following:

The From segment start and Before call end conditions relate to the start and end of each segment and may falsely trigger near a segment break.
The Near and Followed By conditions cannot handle matches that span a segment break, so valid triggers may be missed.
The Not Followed By condition may falsely trigger at a segment break.

Condition Setting	Description
Engine Type	Real-Time Speech Analytics
Project Instance of the Speech and Text Analytics application that supports a different language or Line of Business (LOB).	The project to which the Category to use in this condition belongs. If you do not have Speech Analytics in your deployment, select Predefined categories. If you have Speech Analytics in your deployment, the list includes projects with the real-time categories you have created. In a multi-tenant environment, a Project or Category is available only if the Project or Category is associated with the same tenant as the Recorder Analytics Rule. See the related information to understand how Projects and Categories are associated to tenants in a Multi-Tenant SaaS environment.
Category	The predefined or user-defined real-time category Grouping of calls in Speech Analytics that pertain to specific business issues among all transcribed interactions. to monitor for this condition, based on whether you have Speech Analytics in addition to Real-Time Speech Analytics. If you have Speech Analytics in your deployment, and additional projects with real-time categories, the user-defined categories belonging to the selected project are displayed. If you have only Real-Time Speech Analytics, you can select any of the following predefined real-time categories: Escalations Complaints Customer Positive Sentiment Customer Negative Sentiment Speech driven data masking category, for Sensitive Data Masking
When	The point in the interaction In Speech Analytics, an interaction represents a single part of the contact between one employee and the same customer. In Text Analytics, an interaction is the communication session between one or more employees and the same customer with a unifying contextual element. when the terms and phrases from the selected Category must be detected.
	Anywhere in the call: Matches condition when a word or phrase defined in the selected Category is spoken at any time during the interaction.
	Not followed by: Match condition only when the terms or phrases defined in the first Category selected are detected and the terms or phrases in the second Category are not detected within the specified number of seconds. Project: The project to which the second Category belongs. Category: The second category to compare to the first Category. Within (seconds): The maximum time permitted in seconds, between the terms or phrases detected from the first Category and the words or phrases not detected from the second Category. The default setting is 5 seconds. The minimum setting is 1 second and the maximum setting is 7200 seconds. For example, a time of within 30 seconds means that the condition is met only if the terms or phrases from the first Category are detected , and then the terms or phrases in the second Category are not detected, within the next 30 seconds.
	From segment start: Match condition only when terms or phrases from the selected Category are detected within the specified number of seconds from the beginning of the segment. Within (seconds): The maximum time permitted in seconds from the beginning of the segment for the terms or phrases from the selected Category. The default setting is 5 seconds. The minimum setting is 1 second and the maximum setting is 7200 seconds. RTSA looks at the amount of audio it has received for the current segment when determining if a category match fulfills a From segment start condition. There is never a reference to a timestamp. RTSA determines if the time offset of the category match (that is, where in the audio stream the match was found) is less than the time interval specified in the From segment start condition.
	Before call end: Match condition only when terms or phrases from the selected Category are detected within the specified number of seconds before the end of the segment. Within (seconds): The maximum time permitted, in seconds, from the end of the segment for the terms or phrases from the selected Category. The default setting is 5 seconds. The minimum setting is 1 second and the maximum setting is 7200 seconds.
	Followed by: Match condition only when the terms or phrases defined in the first Category selected are detected within the specified number of seconds after the words or phrases defined in the second Category. Project: The project to which the second Category belongs. Category: The second Category to compare to the first Category. Within (seconds): The maximum time permitted, in seconds, between the terms or phrases detected from the first Category and the words or phrases detected from the second Category. For example, a time of within 30 seconds means that the condition is met only if the terms or phrases from the first Category are spoken, and then the terms or phrases in the second category are also spoken within the next 30 seconds. The default setting is 5 seconds. The minimum setting is 1 second and the maximum setting is 7200 seconds.
	Near: Match condition when any term One to five words that are meaningful to a specific type of business, or phrases that stand out in interactions in Speech and Text Analytics. or phrase defined in the first Category or any term or phrase defined in the second Category are detected within the specified number of seconds of each other. The order in which the words or phrases are detected is irrelevant. This condition is true whenever the terms or phrases defined in either Category are detected within the specified number of seconds of each other. Project: The project to which the second Category belongs. Category: The second Category to compare to the first Category. Within (seconds): The maximum time, in seconds, between the terms or phrases detected from the first Category and the words or phrases detected from the second Category. The default setting is 5 seconds. The minimum setting is 1 second and the maximum setting is 7200 seconds. For example, within 30 seconds means that the condition is met when the terms or phrases from the first Category or from the second Category are detected within 30 seconds of each other.
Side	The channel or the speaker type to match for the condition. All: Match condition when spoken by either employee or customer. For the same condition. the speaker type selected must be valid for the When value. If the When value requires two sets of Categories, you cannot match the condition based on events in both channels. The condition always applies to the same channel. For example, when using the All option: Near: A condition that triggers when "change" is spoken near the phrase "credit card" requires that both terms are spoken by either the employee or the customer. Followed by: A condition that triggers when "account" is spoken, followed by the word "cancel", requires that both terms are spoken by either the employee or the customer. Not followed by: A condition that triggers when "cancel your account" is spoken, but not followed by "help", requires that whoever (customer or employee) speaks the first term does not speak the second term within the given time interval, regardless of what the other person speaks. Employee: Match condition only when spoken by employee. Customer: Match condition only when spoken by customer. For Positive and Negative Customer Sentiment categories, select only Customer.