AutoClassifier 7.1.0.0 Release Notes

This page outlines all of the updates and enhancements in this release. For more information on upgrading, see How to Upgrade AutoClassifier.

April 2025- v7.1.0.0

The following features and updates are new in this release of AutoClassifier:

System Enhancements

Feature

 Description
AI metadata extraction

This release adds the OpenAI information extractor component. The Azure OpenAI information extractor component allows you to use and configure natural language queries from Azure OpenAI and OpenAI LLMs to extract information from your documents. The component can generate suggestions of metadata and questions based on documents that you provide during configuration.
Contextual Metadata Merger This release adds the Contextual Metadata Merger component. This component can refine, filter, and concatenates your metadata to only include relevant information. Additionally, this component can prefix your concatenated metadata to another metadata property, such as the Chunk body property from the document chunker, to enhance document retrieval with more accurate results.
OpenAI and Azure OpenAI embeddings generator components This release adds the OpenAI and Azure OpenAI embeddings generator components. These components allow you to automatically apply vector embeddings to specified properties in your documents. When applied, these embeddings can be used to enhance the accuracy of your search results and provide a better natural-language experience for your users.
RESTful API compatibility

The Restful API Caller component allows you to integrate AutoClassifier with any RESTful APIs (LLMs, AI Cloud Services, etc.) by enabling flexible endpoint configurations, request customizations, and API response management to facilitate efficient data transformation and metadata generation.

For more information, see Apply RESTful API functionality to your content.
AI Visual Interpreter This release adds the AI Visual Interpreter component. This component allows you to use an AI model to analyze visual content, such as images, graphs, and non-structured data, from your documents. When this information is interpreted and extracted, it can be appended as metadata and used to enrich chat bots with context-aware responses. Users will then be able to query their documents for information that is interpreted from non-textual assets.
Document Intelligence enhancements

The Document Intelligence component has been updated to support the following:

  • You can now use the Output format style setting to specify whether you would like your output in text or markdown format.

  • You can now use the Max Pages Per Request setting to divide your documents into segments based on the specified number of pages. This allows the Document Intelligence component to support documents with more than 2000 pages.

Document Chunker enhancements

The Document Chunker component has been updated to support the following:

  • You can now use the Excel Chunking Settings to add additional chunking capabilities for Microsoft Excel files such as splitting by sheets, extracting a specified number of rows as headers, specifying the size of your excel chunks, and specifying the output format.

  • The ChunkPageNumber metadata property has enhanced logic for capturing the respective worksheet number and slide number for Microsoft Excel and PowerPoint files.

  • The output metadata now returns the ChunkBodyForVectorization property, which provides the plain text of ChunkBody property, removing the HTML tags.

  • The output metadata can now include the following metadata, depending on your configuration:

    • ChunkWorkbookName: This property displays the name of your Microsoft Excel workbook.

    • ChunkWorkSheetName: This property displays the name of your Microsoft Excel sheet.

    • ChunkSlideTitle: This property displays the title of the PowerPoint slide.

    • ChunkSlideSectionName: This property displays slide section name of the PowerPoint slide.
Video Indexer enhancements The Video Indexer component has been enhanced to now support extracting subtitles that are included in videos. To use this feature, you must enable the Extract OCR field in the component configuration.
Credential storage enhancements

You can now store your credentials and sensitive information in Azure Key Vault for retrieval by the AutoClassifier scripting component. This prevents sensitive information from being included or exposed in your AutoClassifier scripts.

For more information, see Configuring your Azure key vault storage in the Configure your configuration settings topic.
Scripting enhancements

The following updates are new to the AutoClassifier script component in this release:

  • The AutoClassifier script component can now retrieve the entire list of item properties for a content source in either XML or JSON format. For more information, see Using custom logic.

  • AutoClassifier no longer stores API keys, access keys, or other secrets in Pipeline Stages and configuration files. Now, all credentials are stored in Azure Key Vault storage. For more information, see Configuring your Azure key vault storage in the Configure your configuration settings topic.

  • The script component can not execute or spawn any external processes. Only importing libraries is permitted within the script component.

Bug fixes

  • In some scenarios, the AutoClassifier document chunker component would cause Connectivity Hub crawls to fail. (BAI-17350)

  • When selecting the settings icon in the script component, the Namespaces window displayed in an incorrect location. (BAI-17159)

BREAKING CHANGES

The following are breaking changes that may affect your integration. Please ensure that you carefully review this list before upgrading to AutoClassifier 7.0:

  • When upgrading Connectivity Hub 4.0 to integrate with AutoClassifier 7.0, the URL that is used in the service URL to communicate with AutoClassifier must be updated. For more information, see Configure Your Azure Cognitive Search Target in the Connectivity Hub documentation.

  • The default installation folder is now <program files>\Upland BA Insight\AutoClassifier\. Please follow the install and upgrade instructions to ensure a seamless transition when upgrading to AutoClassifier 7.0.