NICE Uptivity Speech Analytics Overview

You can configure the application to search and index recorded calls using criteria and tags that are important to your organization. When speech analytics is configured with effective criteria and tags, you can use reports to gain insights into interactions in your contact center. In addition to the information in the Uptivity online help, you can refer to a separate PDF guide, Developing an Analytics Program.

Speech analytics must beinstalled on a dedicated server separate from call recording and the NICE Uptivity Web Portal. For general server and environmental requirements, see NICE Uptivity Speech Analytics Requirements. Your NICE Uptivity Sales Engineer will provide customized specifications for this server based on your expected call volume and other factors.

NICE Uptivity Speech Analytics Workflow

When an organization uses search grid-based NICE Uptivity Speech Analytics, the platform uses the workflow described below. This workflow is significantly different from the older, Aurix-based NICE Uptivity Speech Analytics. Throughout the workflow, communication between services uses RabbitMQ. With search grid-based analytics in place:

1. NICE Uptivity records an interaction.
2. CTI CoreThe software component that provides the PBX/ACD integration and makes call recording decisions based on customer-defined recording schedules. sends a message to the Post Interaction Manager (PIM) service that a new recording is complete.
3. PIM informs the TranscoderAn Uptivity service that converts raw files recorded by the system into compressed, formatted files optimized for storage and playback. to transcode the call. If the call was recorded under a schedule that has Speech Analytics set to Yes, PIM tells Transcoder to create a high-quality WAV copy of the recording.
4. Once Transcoder has completed its work and informed PIM, PIM tells the Analytics Manager that a recording is available and ready for analytics processing.
5. Analytics Manager handles communication with the search grid on the NICE Uptivity Speech Analytics server to ensure the file is analyzed and to store the resulting analytics data in the NICE Uptivity database.
6. Once the analytics work is complete, Analytics Manager causes the analytics WAV file to be deleted.

The following table lists the components involved in a NICE Uptivity system that includes speech analytics, and provides additional detail on each component.

Component	Function
NICE Uptivity Server	Records the audio and creates the raw call audio files. Manages the writing of files to storage. Creates call record files in the database. The NICE Uptivity TranscoderAn Uptivity service that converts raw files recorded by the system into compressed, formatted files optimized for storage and playback. service creates two call audio WAV files from the raw files. One file is used for QM and reporting; the other file is used for NICE Uptivity Speech Analytics.
NICE Uptivity Database	Stores call records associated with call audio WAV files as well as post-analytic processing data created by NICE Uptivity Speech Analytics and associated with a call record.
NICE Uptivity Speech Analytics Server(s)	Hosts the Search Grid module which provides processing for NICE Uptivity Speech Analytics. In most cases, also provides storage for WAV files produced specifically for speech processing. These recordings are deleted once analysis is complete.
NICE Uptivity Web Portal	The NICE Uptivity Web Portal is used for analytics reporting, for viewing speech tags while playing calls, and for administering NICE Uptivity Speech Analytics. Call recordings can be filtered using the Tag Cloud Quick Filter. Tags, confidence levels, and phrases appear with call recordings when they are replayed.

Supported Languages

NICE Uptivity Speech Analytics in version 17.3 and later supports multiple languages within a single implementation. The 17.2 release of NICE Uptivity Speech Analytics supports one language at a time. Future releases will support multiple languages within a single implementation. The search grid supports 50+ languages, including (but not limited to):

• Australian, North American, and UK English
• Castilian and Latin American Spanish
• Canadian and European French

Let your NICE Uptivity representative know if you need to use a language other than the languages listed above with your NICE Uptivity Speech Analytics implementation.

Encryption

NICE Uptivity Speech Analytics stereo WAV files are not encrypted. They are permanently deleted once speech processing is complete.

Blackouts

If your organization uses blackouts to protect sensitive data, these blackouts are also applied to the stereo WAV files created for NICE Uptivity Speech Analytics and audio in blacked-out periods is not analyzed. For more information on blackouts, search online help for keyword blackouts or talk to your administrator.

User Permission Settings

Many of the tasks related to NICE Uptivity Speech Analytics are limited by permissions. Depending on the tasks they will perform, NICE Uptivity users must be granted these permissions:

• Allow Analytics View — Allows the user to view NICE Uptivity Speech Analytics data for calls played from the NICE Uptivity Web Portal.
• Allow Analytics Management — Allows the user to manage NICE Uptivity Speech Analytics configuration.
• Allow Viewing Analytics Reports — Allows the user to run and view reports on NICE Uptivity Speech Analytics.

If you need to perform a task and are unable to do so, see your supervisor or NICE Uptivity administrator.

Auditing

Auditing is managed through the NICE Uptivity Web Portal. For more information, search online help for keyword audit log.

Stereo Audio Recordings

Stereo audio recordings improve the performance of NICE Uptivity Speech Analytics, as crosstalk (parties talking at the same time) can be isolated. Within a stereo audio recording, the caller's audio is recorded on one discreet channel for the audio file, and the PBX audio is stored on the other channel. This allows NICE Uptivity Speech Analytics to separate the crosstalk.

Stereo recordings are available on all NICE Uptivity recording integrations with the exception of:

• Avaya DMCCAn API from Avaya that provides the ability to record calls directly from Avaya Communication Manager. DMCC is an acronym for Device Media Call Control. including Single Step Conference, Multiple Registration, and Service Observe.
• ShoreTel TAPIAn acronym for Telephony Application Programming Interface - a Microsoft Windows API that provides computer telephony integration and enables PCs running Microsoft Windows to use telephone services. WAV.
• eOn Service Observe.
• Any recording method using a service observe code via T1.

Save a Second Stereo Audio File

You can request configuration to create and save a second copy of stereo audio files from recordings and save them to another directory. This can be useful if you have a need to provide audio files to another business process. Note that: 

• You are responsible for managing disk space on the second directory. The Archiver does not interact with these files.
• These second copies of audio files do not use File Level Encryption (FLE). They must remain unencrypted in order to be used by another process.
• The second copy of the file is saved as the same file format, audio quality, and size. The only difference from the original is the filename.

Exception Reporting

NICE Uptivity Speech Analytics can identify what is said and, through exception reports, what is not said. Exception reporting can generate a lot of data depending on the number of speech tags and phrases used and whether agents do or do not use the tags. An accurate estimate of the potential size of the data collected is impossible. Therefore, exception reporting is not enabled by default and should be enabled only after consulting with a NICE Uptivity Sales Engineer.

For example, a customer has 1,000 calls per day. The customer uses 30 tags. The difference in data set size with and without exception reporting can be explained in the following way:

• 3 tags are found in each call
• 27 tags are not found in any calls
• Without exception reporting, 3,000 tag records are created. With exception reporting, an additional 27,000 tag records are created for exceptions.

Multi-Site Implementations

NICE Uptivity supports the coordination of multiple speech analytics servers located across geographically-separate sites. An advantage of multi-site implementations is that media files can be processed locally, eliminating the need to send large media files across a WAN to be processed at a different site.

The number of days PAT (phonetic access track) files are retained applies to all Speech Analytics servers in the environment. The PAT retention days setting is configured in the Web Portal.