Release Notes for Mindbreeze InSpire
Version 26.1

Release Highlights

Mindbreeze Insight Workplace

Mindbreeze Insight Workplace serves as a central entry point for accessing existing company data. Due to the seamless connection and analysis of information from different systems, the access to the relevant content and how it is used is greatly simplified.

Relevant company information is consolidated in the Mindbreeze Insight Workplace and made available via a uniform interface. This eliminates the need to switch between individual applications and allows users to use information efficiently and in relation to the context.

Mindbreeze Insight Touchpoint

Insight Touchpoints are a central element of the Insight Workplace. Insight Touchpoints are predefined applications, similar to smartphone apps, that are tailored to specific business requirements. Every interaction in the Insight Workplace is automatically saved as an Insight Touchpoint and is geared toward specific roles, functions, and business objectives.

Instead of individual prompts or generic chat interactions, Insight Touchpoints are preconfigured with defined data sources, retrieval logic, and governance rules. This ensures the consistent, secure, and traceable use of AI in a corporate context.

Mindbreeze Insight Journey

Multiple Insight Touchpoints can be linked to form an Insight Journey – a logical sequence of interactions that build on each other. These workflows are based on real business processes and also support users in efficiently performing complex tasks, such as processing support tickets.

Instead of manually compiling information from different data sources, support staff receive a centralized, context-related overview via combined Insight Touchpoints. Depending on the requirements, individual touchpoints provide recommendations for action, technical documentation, or customer-specific information, such as completed tickets.

Mindbreeze Insight Journeys are automatically created and stored in the Mindbreeze Insight Workplace. They update in real time and access the latest, most reliable data at all times.

Mindbreeze InSpire receives WACA Silver Certificate again

Mindbreeze InSpire has successfully completed the recertification process of the Web Accessibility Certificate Austria (WACA) and has been awarded the WACA Silver Certificate for German and English. During the certification process, it was determined that Mindbreeze InSpire largely meets the WCAG 2.2 – AA success criteria and that the whole content, basic functionality and optional features are accessible to all users.

WACA is an Austrian quality seal that assesses and certifies the accessibility of websites. The website is tested against 56 success criteria that comply with the international W3C guidelines (WCAG 2.2 – AA). The WACA certificate is issued for German and English-language websites in Europe in bronze, silver, and gold levels and is awarded in accordance with EU standard EN 301 549.

Link to the WACA Silver Certificate

New, Extended and Updated

Connection of Data Sources & Indexing

Optimization of the vector index and similarity search for faster and more resource-efficient searches

With the Mindbreeze InSpire 26.1 Release, Mindbreeze customers can find their content even faster while using resources more efficiently and keeping the quality consistent. Thanks to the scalar quantization of sentence embeddings, similarity searches can now be performed more efficiently. GenAI use cases also benefit from this, as retrieval in RAG pipelines is now carried out more quickly.

Link to the documentation

Microsoft SharePoint Online Connector – Upcoming switch to client certificate authentication

Microsoft will discontinue its Azure ACS (Access Control Services) authentication service in April 2026. To respond to this change early on and continue to ensure the functionality of the Microsoft SharePoint Online Connector, the authentication for the connector will be performed with client certificates and a Microsoft Entra ID app for app-only access to Microsoft SharePoint in the future.

Mindbreeze customers will receive information about the authentication changeover in good time via email. If you encounter any problems, the Mindbreeze Support Desk team is available via the ticketing system: https://tickets.mindbreeze.com/

Atlassian Confluence REST Connector – Indexing documents from Atlassian Data Center

With the Mindbreeze InSpire 26.1 Release, the Atlassian Confluence REST Connector now supports the indexing of documents from the Atlassian Confluence Data Center. Mindbreeze customers thereby have the option to index both the Data Center and Atlassian Confluence Cloud and use Mindbreeze InSpire features such as 360-Degree Views and AI Chat to gain new insights from their company data.

Due to the modern design of the connector, the indexing of documents from the cloud and the Data Center is easy to execute. In addition, future maintenance can be carried out without any complications.

Link to the documentation

Applications

Optimized display of Microsoft PowerPoint templates and Office documents

With the Mindbreeze InSpire release, users can now also use the preview function for Microsoft PowerPoint templates with the potx. file format. In addition, compatibility with Microsoft Office documents has been increased to optimize the representation.

Security

Various updates

The Mindbreeze InSpire 26.1 Release includes various security updates for Chromium, Core OS, LibreOffice, and Dell firmware.

Additional Features

Liveness probes provide a more detailed overview of Kubernetes containers

Liveness probes are now automatically available and provide up-to-date status information on individual Kubernetes containers. This gives administrators a detailed overview of the containers and allows them to identify any acute problems in a container at an early stage. In addition, liveness probes check the availability of the GPU, hard disk space, and the current status of the license.

Changelog

• Updated: Chromium to version 144.0.7559.109 (CVE-2025-13630, CVE-2025-13631, CVE-2025-13632, CVE-2025-13633, CVE-2025-13634, CVE-2025-13720, CVE-2025-13721, CVE-2025-13635, CVE-2025-13636, CVE-2025-13637, CVE-2025-13638, CVE-2025-13639, CVE-2025-13640, CVE-2025-14372, CVE-2025-14373, CVE-2025-14765, CVE-2025-14766, CVE-2025-14174, CVE-2026-0628, CVE-2026-0899, CVE-2026-0900, CVE-2026-0901, CVE-2026-0902, CVE-2026-0903, CVE-2026-0904, CVE-2026-0905, CVE-2026-0906, CVE-2026-0907, CVE-2026-0908, CVE-2026-1220).
• Updated: CoreOS to version 43.20260105.3.0 (CVE-2025-9820, CVE-2025-68973, CVE-2025-68972, CVE-2025-9615, CVE-2024-25621, CVE-2025-13601, CVE-2025-15224, CVE-2025-15079, CVE-2025-14819, CVE-2025-14017, CVE-2025-14524, CVE-2025-13034, CVE-2025-15467, CVE-2025-69419, CVE-2025-11187, CVE-2025-66199, CVE-2025-68160, CVE-2025-69418, CVE-2025-69420, CVE-2025-69421, CVE-2026-22795, CVE-2026-22796, CVE-2025-15468, CVE-2025-15469).
• Updated: Dell Firmware (CVE-2024-42446).

• Updated: LibreOffice to version 25.8.
• Added: Liveness probes to Kubernetes to check GPU availability, disk space and license status.
• Added: Preparation for Insight Workplace Persistance Database.
• Added: Custom labels for GPU presence on InSpire Pods.
• Added: Vector index supports scalar quantization (fp16, int8).
• Added: Support for older CUDA versions in RAG.
• Added: Support for Polyfill for XSLT in the Management Center for parsing the config XML.
• Added: New script inverted_index_docinfo.sh to convert dynamic aggregated metadata to static aggregated metadata.
• Added: The property expression “annotation(<property>, "mes:key")” is usable inside the “lookup” to dereference an object.
• Added: Support for Data Center in the Atlassian Confluence REST Connector.
• Fix for: Saved search had to be clicked twice to change the search tabs.
• Fix for: An error occurred when the page was switched when a RAG pipeline/an LLM was tested in the RAG administration.
• Fix for: Jira crawler processed all documents after taking the timezone into account.
• Fix for: Authentication through a credential rather than a clear username/password input is enabled.
• Fix for: GPU driver could not start in an Oracle Cloud Infrastructure.
• Fix for: Client was unable to initialize, when the loading of plugin-config.js failed.
• Fix for: An error occured when  two RAG evaluations with the same metrics were running at the same time.
• Fix for: An issue occured where config syncs in an multi-node setup could trigger the restart of the index (even without configuration changes) under certain conditions. The plugins VocabularySynonymTransformer, ReplacementTransformer and Query Expression Transformer were put back to Java 21.
• Fix for: Newer OpenAI LLMs did not work with Mindbreeze.
• Fix for: A bug with incorrect search constraints occured when tabs were switched that contained a chat widget.
• Fix for: It was not possible to switch tabs in the Settings modal in an Insight Search App.
• Improved: Translations and texts in the RAG UI.
• Improved: Cleaned up the “Generate Embed Code” section in the Insight App Designer by removing unnecessary HTML elements.
• Improved: Charts can now be selected or deselected when they are being used for filtering and they now also highlight which filter is currently active.
• Improved:
• Changed: Element height is only set on the offsetCandidate.
• Changed: Authentification for SharePoint Online Connector is performed with client certificates.