Table of Contents

Features

    Knowledge Base Article Generation  3

        UI Changes for Content Generation from Tickets  3

        Similar Document Comparison  4

    Flows  4

        Workflow Export and Import 4

    Conversation  6

        Status Indicator for Webchat 6

Enhancements

    Conversational AI 7

        Manual Flow Search Reindex for Conversation 2.0 Bots  7

        App Instructions Now Support Hyperflows  7

        Improved RAG Search Accuracy Using Document Titles  8

    Knowledge Base Article Generation  8

        Knowledge Document Count Now Reflects Active Documents Only  8

    AI Workflow Studio  8

        Improved Descriptions and Tooltips for Commonly Used Actions  9

    Connectors  9

        SharePoint support using the File Data Data Source  9

Issues Resolved in the Release  

Features

The following features have been added to the Aisera Gen AI Platform in this release.

Knowledge Base Article Generation

The following features have been added to KB Article Generation in this release.

UI Changes for Content Generation from Tickets

The Bulk tab on the AI Generated Documents window displays the Knowledge Base Articles that were generated from Bulk Tickets to create information Clusters.

The Single tab on the AI Generated Documents window displays the Knowledge Base Articles that were generated from Individual Tickets.

See Bulk and Single Generation.

Similar Document Comparison

Use the Similar Documents window to keep your knowledge articles current. This feature compares existing knowledge articles with newly generated knowledge articles.

This view helps you identify outdated content and review Ai recommended updates with minimal effort.

See Similar Documents for details.

Flows (Workflows and Hyperflows)

The following features have been added to Flows in this release.

Workflow Export and Import

AI Workflow Studio supports a new job-based export and import process for workflows. Resource connections and mappings are handled automatically, and dependencies are configured before an import is finalized.
 

To export Workflows

1. In the Aisera Admin UI, navigate to AI Automation > AI Workflow Studio.
2. Do one of the following:
   • Open a specific workflow, select the More Options menu, then select Export.
     
   • Select one or more workflows from the list view, then select Actions > Export.
3. Navigate to Settings > Jobs, then and wait for the export job to complete.
4. Select the Link to download in the Output field of the completed export job to download the exported workflows

To import Workflows:

1. Switch to the target app and navigate to AI Automation > AI Workflow Studio.
2. From the Workflows List page, open the Import dropdown and select Import.
3. Select the exported workflow file from your local machine. A Pre-Import job will start automatically.
4. Navigate to Settings > Jobs and wait for the Pre-Import job to complete.
5. Navigate back to AI Automation > AI Workflow Studio and select Configure Imported Resources at the top of the page to open the Resource Configuration Mapping Wizard.
6. Review and configure any integration or data source mappings, then select OK to confirm.
7. Monitor the final import job in the Job List page.
8. Once complete, navigate back to AI Automation > AI Workflow Studio to verify the imported workflows.

If no mappings appear in the Resource Configuration Mapping Wizard, select OK to proceed directly to the final import job.

Imported workflows are set to Published status by default. Testing and any additional configuration should be completed before putting them into use.

Conversation

The following features have been added to Conversation in this release.

Status Indicator for Webchat

The Webchat channel now displays real-time connection status notifications to inform bot users when chat connection is interrupted. Bot users will see one of three status indicators at the bottom of the chat window:

• Reconnecting – Displays when the connection is lost and automatic reconnection attempts are in progress, including the current attempt number (for example, "Attempt 4 of 10").
• Back Online – Confirms when the connection has been successfully restored.
• Reconnection Failed – Notifies the user if all reconnection attempts are exhausted, with a Retry option to manually attempt reconnection.

These indicators help users understand why messages may not be sent or received, and when they can resume the conversation.

Enhancements

The following enhancements have been added to the Aisera Gen AI Platform in this release.

Conversational AI

The following enhancements have been made to Conversational AI in this release.

Manual Flow Search Reindex for Conversation 2.0 Bots

For Conversation 2.0 bots, you can now manually trigger a Flow Search indexing directly from the UI when you have to make any changes to the workflows being used by your app.

Running the index manually is useful if you have made changes to Workflow or Hyperflow descriptions or when Workflows or Hyperflows have been added or removed from the Flow Registry, since these changes do not take effect until a reindex is completed. To trigger the reindex, navigate to AI Automation > AI Workflow Studio, click Actions on top right and then click Run Flow Indexing.

Allow approximately 10 minutes for changes to take effect before testing the Workflow or Hyperflow.

App Instructions Now Support Hyperflows

The App Instructions feature now supports Hyperflows in the [Action] field, in addition to standard workflows. When a Hyperflow exists that matches a user Request, it will be identified and returned automatically. If both a Hyperflow and a Workflow exist for the same query, the Hyperflow takes precedence, allowing you to take advantage of more powerful, complex workflows where available.

To configure App Instructions, navigate to LLMs > Prompts Studio in the Aisera Admin UI and select the 
convai_app_instructions_prompt_gpt4_chat_completions prompt.

Select (double-click) the prompt to open the Prompt Playground that allows you to customize and modify the prompt. 

Enter information between the <actions></actions> tags within the code in the System message field.

Improved RAG Search Accuracy Using Document Titles

The Aisera Assistant now includes document titles as part of the RAG private knowledge search, improving the accuracy and contextual relevance of search results.

Knowledge Base Article Generation

The following enhancements have been made to KB Article Generation in this release.

Knowledge Document Count Now Reflects Active Documents Only

The Knowledge Documents count has been updated to reflect only active AI-generated documents, excluding any that have been deleted. Previously, the count did not account for deletions.

With this fix, you now have an accurate, up-to-date count of your active AI-generated documents. Navigate to the SOR > Tickets menu in the Aisera Admin UI to view the AI Generated Documents window that now includes better visibility into your knowledge base articles.

AI Workflow Studio

The following enhancements have been made to AI Workflow Studio in this release.

Improved Descriptions and Tooltips for Commonly Used Actions

The following commonly used Actions have been updated with improved descriptions and tooltips for better clarity:

• Get Due Campaigns
• Get Reminder Eligible Campaigns
• Get Object
• Put Object
• Generic Event Action

Connectors

The following enhancements have been made to Connectors in this release.

SharePoint support using the File Data Data Source

The File Data Data Source now supports SharePoint integrations, enabling ingestion of CSV and JSON files directly from SharePoint Document Libraries. To configure a File Data Data Source for SharePoint ingestion:

1. Navigate to Settings > Data Sources.
2. Select + New Data Source.
3. Select the File Data Data Source, then select Next.
4. Under Files select Cloud Files.
5. Select your SharePoint Integration in the Integration dropdown menu.

Three new optional fields are available:

• Sites: Specify one or more SharePoint sites to ingest from. If left empty, files are ingested from the default root site.
• Libraries: Specify one or more libraries to ingest from. If left empty, files are ingested from all libraries.
• Folders: Specify one or more folders to ingest from, including all subfolders. If left empty, the entire file system of the selected libraries or sites is searched.

Each field accepts a newline-separated list of values. More specific inputs result in faster processing.

Field mapping must be configured to properly process CSV and JSON files.

Issues Resolved in the Release