The capabilities introduced in V‑Spark version 3.0 can be organized into several general areas of improvement:

1. Navigation

   1. All pages under Settings will have the same breadcrumb-style navigation as the Dashboard.

   2. Folders and Applications are now displayed on separate pages.

   3. The Application Editor is now accessible from the Applications Page.

   4. Application reprocessing is now controlled directly from the Applications page.

2. Dashboard Graphs & Statistics

   1. Overview and Applications graphs can be set to display historical data (data that is older than thirty days).

   2. Overview Call Statistics table now displays monthly averages.

   3. Call Statistics values will not be updated for data uploaded with historical date times.

   4. Application graphs (both Overview and Daily Chart) now display the number of calls that matched each category.

3. Dashboard Search

   1. Time-based ad-hoc searches can now be conducted.

      1. Proper format: search phrase followed by ~s or ~e (to indicate the beginning or end of the call), a greater/less than symbol (>/<), and the number of seconds (to indicate which section of the call to search).

      2. For example, "thank you < 30" will find all calls where "thank you" was said during the last 30 seconds of the call.

      3. Time-based searches terms can be excluded as well, using a minus symbol (-).

      4. Time-based search will only be enabled for data uploaded after updating to version 3.0.

   2. New ad-hoc search capabilities have been added so that users can use (^) and ($) to search for phrases at the very beginning or end of a speaker turn, respectively.

      Note: These searches will only be enabled for data uploaded after updating to version 3.0.

   3. Application search has been improved so that users can view results for multiple application categories at once, and perform Boolean AND/OR searches on these selections.

4. File Details Page

   1. Application scores are fully integrated into the File Details page. Phrases matching each application category can be highlighted individually.

   2. Silent portions of the call are now displayed as part of the transcript so users can see where silence occurs and for how long. Users will also be able to skip these silent portions when replaying audio.

   3. More efficient display of File Details menu:

      1. File names are now displayed above the audio player.

      2. Controls for turning emotion visualization and silence skipping on/off have been moved above the call metadata for easier access.

   4. The name of the file being viewed will appear within the browser tab for easier navigation between tabs.

5. Enhanced Application Support

   1. Key Highlights

      1. Version 3 now includes the following application templates:

         1. Agent Scorecard (improved from version 2)

         2. Call Drivers (improved from version 2, formerly Voice of the Customer)

         3. Customer Experience (includes analyses for first call resolution, churn detection, and customer sentiment)

      2. Multi-level applications now support more in-depth analysis with a category hierarchy of up to 5 categories deep by default.

         1. Currently configured applications will be converted to the new underlying JSON structure.

         2. Application configuration files with the old JSON structure are no longer supported and cannot be uploaded to custom applications.

         3. Ad-hoc application search for data uploaded before updating to version 3 will only work for top-level categories.

      3. New applications can be created by copying any existing application within the user’s available organizations.

   2. Editing

      1. New application search capabilities have been added for more flexible analysis. More detailed explanations of each search character are available in the "V‑Spark Application Development Guide".

         1. Underscore (_) now indicates 0 or more words, rather than 1 or more words.

         2. Boolean searches find files that contain both (&) or at least one (|) search phrase.

         3. Distance-based search (~n) finds files containing phrases that appear up to n words apart. Turn-based search (~t) finds files containing phrases that appear within a single speaker turn.

         4. Wildcards for single characters (?) and 0+ characters (*).

         5. Location-based searches find files containing phrases at the very beginning (^) or very end ($) of the speaker turn.

         6. Searches can be conducted across speaker turns using ( -->), e.g. "how can i help you --> billing" will match any calls where Turn A contains "how can i help you" and Turn B contains "billing".

      2. Custom metadata values may be added to applications as include and exclude filters so that users can better tailor analysis to their particular company.

      3. Simultaneous editing of applications by multiple people is now supported.

      4. All application phrases added using the Application Editor, phrase list upload, or JSON config upload will be validated.

         1. Attempts to add invalid phrases will generate an error message.

         2. Invalid phrases added using the Application Editor will not be uploaded. Phrases containing errors in syntax will be uploaded, but corrected to the proper syntax.

         3. Invalid phrases added using the phrase list upload will not be uploaded, but all remaining phrases in the phrase list file will.

         4. Invalid phrases added using a JSON config file will generate an error message listing the first invalid phrase and no portion of the new config file will be uploaded.

      5. Any changes made to an application will be saved automatically as soon as they are entered, rather than at the end of each session.

   3. Scoring

      1. Application scores can now be exported for downstream analytics more easily.

         1. Application scores for custom date ranges, rather than a single day can be exported from the Agents View.

         2. Application scores will automatically be included in JSON transcripts.

         3. Reprocessed scores will be available in JSON transcripts downloaded from the File Details page, but not via bulk download from the Settings page.

      2. Users can select a default scoring type for applications (either Hit/Miss or Coverage) to be displayed on the Dashboard graphs, Agents View, and Summary View).

      3. Application scoring has changed. Key changes have been summarized below. More detailed information can be found in "V‑Spark 3.5.0 Application Development Guide".

         1. Categories at the lowest level of an application (leaf-level categories) will receive a Count score that is equal to the number of phrases that were matched for that category.

         2. Upper-level categories will receive both a Coverage score (formerly Average) and a Hit/Miss score (formerly Total).

         3. The default value for existing applications will automatically be set to Hit/Miss. Users can switch this to Coverage, but the scores displayed will be equivalent to the old Average scores. For new Coverage scores, users will have to reprocess the application.

         4. Exclude phrases in lower-level categories will only cancel out other phrase matches, rather than entire calls.

6. Application Queue

   1. The application queue now includes more detailed status information and a timestamp indicating the time a reprocessing request was received or a folder was unlinked from an application.

   2. The application queue will appear immediately after submitting a reprocessing/delete request.

7. Trending

   1. Trending Editor functionality has been brought in line with Application Editor functionality, with (-) being used to designate exclude phrases, rather than a separate tab.

   2. Users will be able to view a roll-up of trending phrases for custom date ranges, rather than a single day.

8. Logging

   1. Detailed logs of user activity and system status can now be viewed by Admin users, including the following information:

      1. Creation, deletion, and updates to companies, organizations, folders, and applications

      2. Approval of new user accounts as well as updates or deletion of existing user accounts

      3. Reprocessing of applications

      4. Uploading of audio/text files for initial transcription User logins and logouts

   2. Users with Admin and Manager permissions can access a log of transcription status for individual audio/text files from the Settings/Folders page.

9. Upload/Download Improvements

   1. Bulk downloads of all file types can now be done for a custom date range.

   2. MP3 files can now be downloaded in bulk from the Settings/Folders page.

   3. JSON representing text data sources such as chat and email can be consumed without audio.

   4. The system has been optimized for greater overall throughput of transcription processing.

10. Fixes

    1. Emotion metadata filters in applications no longer trigger a reprocessing error.

    2. More consistent highlighting in Dashboard search snippet preview and File Details page.

    3. Trending phrases with equivalent frequency counts are now displayed in alphabetical order.

11. For Developers

    1. V‑Spark clients who have S3 configured can now pull zipped input directly from S3 via the Voci API.

    2. Backend logging has been improved to rotate logs daily and delete logs older than 30 days.

1. Dashboard Displays

   1. Values in the Dashboard Overview Call Statistics table will not be updated for data uploaded with historical datetimes.

   2. Files uploaded to a folder that is linked to a disabled application will still appear in the Files View on the application dashboard.

   3. Files uploaded to a folder before an application has been reprocessed will still appear in the Files View on the application dashboard.

2. Search

   1. Time-based search

      1. Searches are turn-based and may return results for phrases outside of the specified timeframe, but within the same speaker turn that starts or ends within that timeframe.

      2. Search snippets in the Files View may display results outside the specified time interval.

   2. The (-) operator for excluding a search tag in ad-hoc search is not supported for file names.

   3. File Details Highlighting Exceptions* (in order of severity):

      1. Approximate searches using (~a) will not be highlighted.

      2. Searches using (~n) may highlight phrases with the specified number of words in between each of the search terms, rather than between all of the search terms. For example, "get help phone ~4" may highlight phrases with 4 words between "get" and "help" and another 4 words between "help" and "phone"

      3. Searches using (~t) and (~n) will only be highlighted if the words matched appear in the order specified in the search tag.

      4. Highlighting may be incomplete for matches marked with emotion

      5. Highlighting may be incomplete for matches with intervening silence tags greater than 2 seconds.

      6. Matches for "@*" (such as when searching for email addresses) will only be highlighted until the first "." in the email address. Regular expression searches may be used to circumvent this issue.

      7. Excluded application search phrases may be highlighted on the File Details page.

      Note: Note that these exceptions only affect highlighting, and that all search results using the above characters will still be accurate.

3. Other

   1. The Trending View may take some time to fully load, especially for longer date ranges.

   2. Users playing longer audio files via Firefox may encounter playback/syncing issues when jumping to different sections of the transcript. This is an external issue caused by the Firefox browser. Users who encounter this issue can simply wait for the browser to finish buffering the file before resuming their search.