Documentation

Releases and Release Notes

Version 1.23 (October 2026)

  • Alfresco Connector

    • Small adjustments to ACL field calculation

  • Connector Framework

    • Race condition which sometimes caused checksums being wrongly calculated which gave rise to reindexing documents

  • Document Previews

    • Solved small issue which caused memory consumptions going up

  • Query Pipelines

    • Improved query logging capabilities

    • Added means to stop a running query

    • Added offline keyword extraction capabilities

    • Added new result deduplication stage

    • Changed standard model for cross encoder reranker

    • Adjusted issues within the expert search

    • API Access changes

      • Query pipelines for public search do not allow for API access.

      • Pipelines with API access tokens or OAuth authentication do not allow for public search access without API token anymore.

        In turn the API endpoint api/v1/websiteSearch is only available in the following scenario

        • The search interface references one or multiple query pipelines

        • The access configuration of each query pipeline must be set to “Public Search”

      • If you use public search and our query pipelines, the upgrade steps are Version 1.23 Upgrade Steps

Version 1.22 (September 2026)

  • Added OpenProject Connector

  • Added means to index transformed HTML alongside with text snippets

  • Admin experience

    • Revised admin experience

    • Introduced optional connector labels which can be used to distinguish multiple instances of the same connector family

  • Search Engines

    • OpenSearch and Elasticsearch revised standard settings

  • AI and Retrieval Pipelines

    • New AI pipeline step for automated evidence collection solving complex agentic retrieval tasks

    • New reranker step which utilizes the shipped reranker endpoint

Version 1.21 (August 2026)

  • Added XWiki Server Connector

  • Added Magnolia CMS Connector

  • Changes to Confluence Data Center and Server Connector:

    • Filtering for spaces becomes more efficient as these are not stored in the database

    • Added crawl option to only expand groups for group members which occur in the page and space permissions. This allows for handling scenarios where Atlassian Crowd synchronizes large scale Active Directory or LDAP installations with Confluence’s group management.

  • Changes to Microsoft Search and Microsoft Copilot Indexing

    • Improved multithreading for principal ingestion

    • Added handling of new Microsoft-side API I issue “ParentExternalGroupMissingSync”, “The parent external group where member is being added has not been synced yet.”

Version 1.20 (July 2026)

  • Added OpenText Content Server Connector

  • Added Qdrant as new search engine

  • Added URL Encoding for metadata fields

Version 1.19 (June 2026)

  • Added synonym management

    • Automatic LLM based synonym extraction as part of the content transformation pipeline

    • Synonym management dialog

    • Synonym query transformation step

  • Improved UX for query processing (automatic reloads in query logging mode)

  • SharePoint Connector

    • Bug fixes for Umlaut issues for the SharePoint on premise connector

    • Added further means for SharePoint Site autodiscovery, based on search refiners, search for sts site and manual lists

  • SharePoint Online Connector

    • Bug fixes for Umlaut and token refresh issues for the SharePoint Online connector

Version 1.18 (May 2026)

  • Added VLLM support

  • Improved LM model configurations

  • Fixed race condition for crawl initialization

  • Adjusted default memory parameters to 4GB

  • Introduced standalone document previewer

  • Introduced Agentic Retrieval for AI and query pipelines

  • Confluence Cloud: introduced usage of new API endpoint for attachment download

Version 1.17 (April 2026)

  • Significant performance improvement to the SharePoint Online connector

  • Added optional filter configuration for Dynamics to only index resolved incidents

  • Improvements towards the SharePoint on premises connector

  • Extended preview generation and documentation

Version 1.16 (March 2026)

  • Added new Alfresco Connector

  • Extended Squirro API management to “Version 2”

  • Extended Slack connector towards latest Slack API changes

  • Added new Exchange Online Connector

  • Improved stability of text extraction

Version 1.15 (February 2026)

  • Introduced asynchronous AI pipelines

  • Introduced document classification content transformer.
    Uses LLM to extract

    • document scope

    • document type

    • document classifications

    • company and organization names

    • person names

  • Added document translation content transformer. Uses an LLM to automatically translate a document to a default language

  • Extended search schemas for Elasticsearch, Azure AI Search, Apache Solr, OpenSearch, Weaviate towards new metadata fields
    For existing installations, you need to create the new fields using the modify schema functionality (see Index Creation and Management )

  • Improved Open WebUI Support

    • Added detailed activity messages

    • Fixed blocking calls

  • File share connector
    Added support for Linux ACLs

Version 1.14 (January 2026)

  • Added detailed on-demand query execution history and workflow analysis tools

  • Added new RAG workflow stages

  • Improved RAG wizards

  • Added extensive support for on-premise LLM models

  • Dynamics connector

    • Added accounts as metadata to relevant objects

    • Added tasks as content type

    • Changed last modification date computation for incidents, opportunities and leads being based also on the record’s modification date and not only the record interactions

Version 1.13 (December 2025)

Version 1.12 (November 2025)

  • Confluence Connector

    • Improved error handling for occasional glitches in Confluence Version 9 and 10

    • Extended the Confluence on-premises connector to support page rendering in case contents from dynamic macros are needed

  • Content Processing Pipelines

    • Added new text extraction stage which allows for extracting contents from complex HTML elements, such as aura-card macros in Confluence

  • Query Pipelines

    • Added new configuration experience for query pipelines

    • Added deep research mechanisms into query pipelines

    • Added multiple new query processing stages / steps to handle complex reasoning within query pipelines

    • Improved Elasticsearch, Solr and Opensearch query behaviors

  • Query APIs

    • New query APIs for third party integrations with RheinInsights Retrieval Suite

  • Microsoft Search and Copilot Connectors

    • Error handling towards missing user identities in Entra Ids for our Copilot Connectors

  • Squirro

    • Improved metadata handling for our Squirro connectors

Version 1.11 (October 2025)

  • Zendesk Connector

  • Azure Blob Storage Connector

  • Azure File Storage Connector

  • Search Engine / Search Federator

  • Query APIs

    • Added API for direct query pipeline access

    • Moved query side rate limiting into query pipeline

  • SharePoint On-Premises Connector

    • Improvements to SharePoint on-premises connector in case not the right permissions are granted.

  • Confluence Cloud Connector

    • Upgraded Confluence Cloud towards new REST APIs

  • Jira Cloud Connector

    • Upgraded Jira Cloud towards new REST APIs

Version 1.10 (September 2025)

Teams, Slack, Bot and Agent to Agent Integrations

Version 1.9 (August 2025)

Version 1.8 (July 2025)

  • Added Microsoft Teams connector Microsoft Teams Connector

  • Added SQL databases as target system SQL Data Target - Indexing Configuration

  • Added hybrid search to Azure AI Search and Elasticsearch

  • Improved Query Pipeline Configuration Wizard Search Experience and Query Pipelines

  • Improvements to Squirro backend Squirro Insight Engine - Indexing Configuration

  • Fixed an issue during setup of query pipelines, when admin UI authentication is secured with BASIC auth.

  • Fixed an issue when running the Retrieval Suite under Linux or MacOS.

  • Fixed a rare race condition where two starting crawls might interfere with each other.

  • Added update check mechanism and a setting to disable the checks.

  • Improved RAG answering based on Ollama models.

  • Added an option to disable ACLs. This is not recommended, but needed for some customer scenarios.

Version 1.7 (May 2025)

Version 1.6 (April 2025)

Version 1.5 (March 2025)

Version 1.4 (January 2025)

Version 1.3 (November 2024)

Version 1.0 - 1.2 (September 2024)

  • Initial releases