I would like to propose an improvement regarding the automatic launch of the Teams policy wizard when accessing the Teams admin center. Recently, one of our administrators encountered the wizard upon login, and naturally followed the steps presented. Unfortunately, this resulted in a configuration change that impacted over 10,000 users, who suddenly lost their ability to make calls or hold meetings. because the wizard automatically modifies the configurations of Teams global policies.

Given the significant impact that such changes can have on a large user base, I would suggest that the Teams policy wizard not automatically appear upon admin login. Instead, perhaps it could provide a clearer prompt about the consequences of proceeding with the wizard.

This change could help prevent accidental misconfigurations and reduce the risk of widespread service interruptions for Teams users.

Thank you for considering this feedback.

​I would like to propose an improvement regarding the automatic launch of the Teams policy wizard when accessing the Teams admin center. Recently, one of our administrators encountered the wizard upon login, and naturally followed the steps presented. Unfortunately, this resulted in a configuration change that impacted over 10,000 users, who suddenly lost their ability to make calls or hold meetings. because the wizard automatically modifies the configurations of Teams global policies.Given the significant impact that such changes can have on a large user base, I would suggest that the Teams policy wizard not automatically appear upon admin login. Instead, perhaps it could provide a clearer prompt about the consequences of proceeding with the wizard.This change could help prevent accidental misconfigurations and reduce the risk of widespread service interruptions for Teams users.Thank you for considering this feedback.  Read More

​

Announcement below.  Wondering if this will have impact on a partner’s CSP customers or will the current CSP credentials be suffice?

Starting 15 October 2024, we will require users to use multifactor authentication (MFA) to sign into the Azure portal, Microsoft Entra admin center, and Intune admin center. To ensure your users maintain access, you’ll need to enable MFA by 15 October 2024.

If you can’t enable MFA for your users by that date, you’ll need to apply to postpone the enforcement date. If you don’t, your users will be required to set up MFA.

Action required

To identify which users are signing into Azure with and without MFA, refer to our documentation.
To ensure your users can access the Azure portal, Microsoft Entra admin center, and Intune admin center, enable MFA for your users by 15 October 2024.
If you can’t enable MFA by 15 October 2024, apply to postpone the enforcement date.

​Announcement below.  Wondering if this will have impact on a partner’s CSP customers or will the current CSP credentials be suffice?
 
Starting 15 October 2024, we will require users to use multifactor authentication (MFA) to sign into the Azure portal, Microsoft Entra admin center, and Intune admin center. To ensure your users maintain access, you’ll need to enable MFA by 15 October 2024.If you can’t enable MFA for your users by that date, you’ll need to apply to postpone the enforcement date. If you don’t, your users will be required to set up MFA.
Action required

To identify which users are signing into Azure with and without MFA, refer to our documentation.
To ensure your users can access the Azure portal, Microsoft Entra admin center, and Intune admin center, enable MFA for your users by 15 October 2024.
If you can’t enable MFA by 15 October 2024, apply to postpone the enforcement date.  Read More

​

Introducing the Azure Multimodal AI & LLM Solution Accelerator

The Azure Multimodal AI & LLM Processing Accelerator is your one-stop-shop for all AI+LLM processing use cases like content summarization, data extraction, classification and enrichment. This single accelerator supports all types of input data (text, documents, audio, image, video etc) and combines the best of Azure AI Services (e.g. Document Intelligence & AI Speech) and Large Language Models (such as the LLMs available in Azure OpenAI and Azure AI Studio) to achieve accurate, reliable and scalable automation of tasks. Best of all, it enables development teams to build and maintain all of their applications from a single codebase, allowing you to deliver use cases to production much faster and far more maintainable than if you used 5-10 different accelerators (that each only focus on specific use cases or datatypes).

Get started with the Github Repository: https://github.com/Azure/multimodal-ai-llm-processing-accelerator

This accelerator implements many of the traditional machine learning techniques that are lacking in existing GenAI applications to enable true automation of backend processing tasks, including the merging of confidence scores and metadata from Azure AI services with raw LLM outputs. This makes it possible to automatically identify and accept confident/reliable results, while automatically escalating the less reliable results to human review instead of needing to review all LLM outputs to ensure their accuracy. This will help your organization transition to true automation instead of the current state-of-play where every GenAI application is a chatbot (since every LLM result needs to be reviewed).

The accelerator also resolves many of the shortcomings of existing demo applications and code samples, including native support for the full range of real-world integration options (HTTP API, CosmosDB, SQL, Blob, Event grid, AI Search, Fabric and more), a selection of pre-built pipeline templates to allow developer teams to get started immediately (for use cases like document classification & extraction, customer contact centre processing & analytics, content summarization etc), a demo web app for showcasing the prebuilt pipelines or the customer’s custom solutions to non-technical stakeholders, and comes ready with infrastructure templates for full deployment to Azure in 7 minutes flat.

Form Field Extraction Pipeline – Example Output

An example result from the Form Extraction pipeline, which returns confidence scores and bounding boxes with each extracted value. This pipeline then automatically flags the result for manual review if any of the extracted values are below a given confidence threshold

Key Features

Pre-built processing pipelines included: The solution comes with a number of pre-built processing pipelines that can be easily customized and deployed straight to production, such as document processing, text summarization, contact center call analysis and more.
Azure Function host: The solution uses Azure Function App as the pipeline host, offering scalability, cost-efficiency and a full suite of integration options out of the box. Azure Functions also make it easy to configure rate and concurrency limits and concurrency that are crucial when dealing with large-scale production deployments (e.g. to ensure LLM endpoints are not overloaded when a batch process is triggered).
Front-end demo app included: A simple demo web app makes it easy to test backend APIs through a UI, and to share solutions and collaborate with non-technical users and business stakeholders.
Data converters and processors: Many of the core components required for multimodal processing pipelines are included, such as Azure Document Intelligence, Azure AI Speech, Azure OpenAI and more. These help you easily convert your data and Azure AI API responses into the best format for consumption by LLMs, and the pipelines are built in an open and extensible way so that you can easily incorporate custom-built or external pipeline components.
Enriched outputs & confidence scores: A number of components are included for merging the outputs of the pre-processing steps with those from the LLM. For example, merging the confidence scores, bounding boxes & writing styles from Document Intelligence with the values returned by the LLM. This allows for reliable automation of tasks instead of having to trust that the LLM is correct (or reviewing every result).
Data validation & intermediate outputs: All pipelines are built to validate that results conform to the required schema and return not just the final result but all intermediate outputs. This lets you reuse the data (E.g. the raw Azure AI Speech transcription) for other downstream tasks to avoid paying for the processing of the same input data multiple times, while also providing more useful information to the end client.
Powerful and flexible: The application is built in a way that supports both simple and complicated pipelines. This allows you to build pipelines for all of your use cases without needing to start from scratch with a new code base when you need to something a little more complex (a common occurrence with many GenAI frameworks and accelerators). A single deployment of this accelerator can support all of your backend processing pipelines, from proof of concepts to business-critical applications.
Infrastructure-as-code: An Azure Bicep template and instructions for local & cloud deployment are included, guiding the customization process and enabling immediate deployment to Azure.

Process Flow & Solution Architecture

The solution is built to support all types of data and easily integrate into all types of real-world production systems.

Call centre analysis: Transcribe and diarize call centre audio with Azure AI Speech, then use Azure OpenAI to classify the call type, summarize the topics and themes in the call, analyze the sentiment of the customer, and ensure the customer service agent complied with standard procedures (e.g. following the appropriate script, outlining the privacy policy and sending the customer a Product Disclosure Statement).

Document processing: Ingest PDFs, Word documents and scanned images, extract the raw text content with Document Intelligence, then use Azure OpenAI to classify the document by type, extract key fields (e.g. contact information, document ID numbers), classify whether the document was stamped and signed, and return the result in a structured format.

Insurance claim processing: Process all emails and documents in long email chains. Use Azure Document Intelligence to extract information from the attachments, then use Azure OpenAI to generate a timeline of key events in the conversation, determine whether all required documents have been submitted, summarize the current state of the claim, and determine the next-best-action (e.g. auto-respond asking for more information, or escalate to human review for processing).

Customer email processing: Classify incoming emails into categories, summarizing their content, determining the sender’s sentiment, and triage into a severity category for human processing.

Background & Problem Statement

The promise (and challenges) of using LLMs for task automation

Most organizations have a huge number of simple and tasks and processes that consume large amounts of time and energy. These could be things like classifying and extracting information from documents, summarizing and triaging customer emails, or transcribing and running compliance tasks on contact centre call recordings. While some of these tasks can be automated with existing tools and AI approaches, they often require a lot of up-front investment to fully configure, train and customize in order to have a reliable, working solution. They can also be perform poorly when dealing with input data that is slightly different than expected, and may never be the right fit for scenarios that require the solution to be flexible or adaptable.

On the other hand, Large Language Models have emerged as a powerful and general-purpose approach that is able to handle these complex and varied situations. And more recently, with the move from text-only models to multimodal models that can incorporate text, audio and video, they are a powerful tool that we can use to automate a wide variety of everyday tasks. But while LLMs are powerful and flexible, they have their own shortcomings when it comes to providing precise and reliable outputs, and they too can be sensitive to the quality of raw and unprocessed input data.

The challenge for AI systems: Reliability

However, the biggest problems with LLMs is their lack of reliability. Unlike almost every other type of AI model, LLMs do not give reliable or usable confidence scores with their outputs (primarily due to their token-in-token-out architecture). And if an LLM is asked to provide a rating of the reliability of their response, research has shown that they can be unreliable and uncalibrated, especially for single predictions (1, 2, 3). The end result is that the only way to be confident in the results of an LLM is to review every output and turn everything into a chatbot – not exactly the automation that we hoped for.

For decades we have been using machine learning and AI models to automate tasks. In most cases where these models are used to assist in human processes, confidence scores are used to help decide whether an AI prediction can be accepted automatically, or whether it needs to be reviewed (usually referred to as ‘human-in-the-loop’ processing). The idea of this approach is to use the model confidence scores and other metadata to let the model solve the majority of the easy and routine cases while escalating all of the difficult and challenging cases (where the model has lower confidence scores) to a human. With this approach, it is common to automate 80+% of tasks with extremely high accuracy and reliability, while ensuring that the most challenging cases are always reviewed by a human.

For data processing tasks like summarization, data extraction and classification, LLMs will usually give strong results during a proof-of-concept. But most development teams will discover that it is challenging or even impossible to move from a good-enough solution to a reliable production-ready solution due to the fact that their inner workings are largely a ‘black box’. And unlike traditional machine learning pipelines, there are only a handful of things that can be tuned (such as changing the model or tweaking the prompt) to improve performance, with changes usually having far-reaching effects that change the performance across the entire dataset – a truly frustrating experience.

Unlocking the best of both worlds: Combining domain-specific AI models with LLMs

By combining the consistency, reliability and rich outputs of domain-specific AI models (such as Azure Document Intelligence, Azure AI Speech etc) with the general knowledge and flexibility of LLMs, it is possible to build solutions that are accurate, reliable, fast and cost-effective. Most importantly, it is possible to create systems that truly automate tasks and without the need for manual review of every result.

In a recent customer project that involved extracting Order IDs from scanned PDFs and phone images, we used a number of these techniques to increase the performance of GPT-4o-alone from ~60% to near-perfect accuracy:

Giving GPT-4o the image alone resulted in an overall recall of 60% of the order IDs, and it was impossible to know whether the results for a single PDF could be trusted.
Adding Document Intelligence text extraction to the pipeline meant the GPT-4o could use analyze both the image and extracted text. This increased the overall recall to over 80%.
Many images were rotated incorrectly, and GPT-4o performed poorly on these files (50% lower performance due to rotation alone). Document Intelligence returns a page rotation value in its response, and using this to correct those images prior to processing by GPT-4o drastically improved performance on those images – from 50% to over 80%.
Finally, by cross-referencing the order IDs returned by GPT-4o against the Document Intelligence result, we could assign a confidence score to each of the extracted IDs. With these confidence scores, we were able to tune and set confidence thresholds based on the type of image, resulting in 100% recall on 87% of the documents in the dataset, with the final 13% of preliminary results being automatically flagged for human review.

At the conclusion of this project, our customer was able to deploy the solution and automate the majority of their processing workload with confidence, knowing that any cases that were too challenging for the LLM would automatically be escalated for review. Reviews can now be completed in a fraction of the time thanks to the additional metadata returned with each result.

Future Roadmap, FAQs & More

For more information on the future roadmap, FAQs and more, head over to the Github repository. If you have questions about the repository, have feature requests or bug reports, or want to share your experience with the accelerator, please submit your feedback using the Github Issues section.

I hope you enjoy the accelerator and it helps you create value in your organization!

​Microsoft Tech Community – Latest Blogs –Read More 

Hi,

I just imported EASM data to Sentinel, so we can create some analytic rules based on EASM data.

I’m now thinking on which use cases are interesting to create alerts.

Anyone has already followed this path and has some experience of what kind of alerts make sense based on EASM data?

Thanks

​Hi, I just imported EASM data to Sentinel, so we can create some analytic rules based on EASM data.I’m now thinking on which use cases are interesting to create alerts.Anyone has already followed this path and has some experience of what kind of alerts make sense based on EASM data? Thanks  Read More

​

My boss created an Excel data model before I joined the company. He created the data model and queried the data from files within his personal drive. Now that I am here, he wants me to have access to these files and has moved them into a shared folder. The thing is, all of the queries now have the wrong file path. Is the only way to fix this to manually update the file path within the Source step of the power query editor? If there is an easier way, please let me know!

​My boss created an Excel data model before I joined the company. He created the data model and queried the data from files within his personal drive. Now that I am here, he wants me to have access to these files and has moved them into a shared folder. The thing is, all of the queries now have the wrong file path. Is the only way to fix this to manually update the file path within the Source step of the power query editor? If there is an easier way, please let me know!  Read More

​

Windows 11, version 24H2, also known as the Windows 11 2024 Update, is now available through Windows Server Update Services (WSUS) and Windows Update for Business. You can also download it from the Microsoft 365 admin center[1], Software Download Service (via Installation Assistant, the media creation tool, or ISO), and Visual Studio Subscriptions.

Today marks the start of 36 months of support for Enterprise and Education editions of Windows 11, version 24H2[2]. We recommend that you begin a targeted deployment in your organization now to validate that your apps, devices, and infrastructure work as expected with the new release. To help you plan, this post outlines some of the features and enhancements that help you power exceptional experiences while helping secure your corporate data, apps, and people on any device.

New in this update

Windows 11, version 24H2 includes all the features and capabilities delivered as part of continuous innovation to Windows 11, now enabled by default. These include:

Windows Local Administrator Password Solution (LAPS) policy improvements and new automatic account management feature
Personal Data Encryption (PDE) so that known Windows folders (Documents, Desktop, and Pictures) are protected using user authenticated encryption
App Control for Business (formerly Windows Defender Application Control) to better protect your digital property from malicious code
Windows protected print mode, eliminating the need to rely on third-party software installers for Morpia certified printers
Local Security Authority (LSA) protection to help protect against the theft of secrets and credentials used for logon
Support for Wi-Fi 7
Bluetooth® LE audio support for assistive devices
New controls to help manage which apps have access to the list of Wi-Fi networks around you
Rust in the Windows kernel
SHA-3 support

With Windows 11, version 24H2, we’ve also introduced numerous Server Message Block (SMB) protocol changes. These include firewall rule changes, support for blocking NTLM, dialect management, alternative network port connections, SMB over QUIC (an alternative to TCP and RDMA), and changes to SMB signing and encryption.

This scoped release also features enhancements designed to improve your overall experience with Windows 11, such as:

A scrollable list (vs. single pane) of Quick Settings in the system tray that can be arranged in the way that makes sense for how you work best
Text labels to the cut, copy, rename, share, and delete actions available at the top of the File Explorer context menu
Support for creating 7-zip and TAR archives
Energy saver, an easy way to extend battery life and reduce energy use that can be configured to run automatically or toggled on and off manually via Quick Settings

Extended adaptive brightness control[3] on laptops and 2-in-1s even when they are plugged in

Expanded availability of Voice Clarity, which cancels echo, suppresses background noise, and reduces reverberation in real-time

We’ve also added new capabilities to control audio presets, ambient sounds, and more for Bluetooth® LE Audio. We made it so that people can adjust audio presets for their hearing aids from Settings or Quick Settings.

For more information on what else is new—including a list of features no longer under temporary enterprise control and features exclusive to Copilot+ PCs—check out What’s new in Windows 11, version 24H2.

You can also take a closer look with this special episode of Microsoft Mechanics:

Updated deployment and management tools

Many of the deployment, security, and management tools you rely on have been updated or refreshed and are also available today! This includes:

Windows 11, version 24H2 Security Baseline – Download Microsoft-recommended configuration settings and explanations of their security impact.
Administrative Templates (.admx) for Windows 11, version 24H2 – The most popular request! While natively accessible via the C:WindowsPolicyDefinitions folder in Windows, you can download the ADMX separately and use them to populate policy settings.
Group Policy settings reference spreadsheet for Windows 11, version 24H2 – Download a spreadsheet of policy settings for computer and user configurations that are included in the ADMX files for Windows 11, version 24H2.
Remote Server Administration Tools (RSAT) for Windows 11 – Manage Windows Server roles and features from a Windows 11 device. To add RSAT, which are included as Features on Demand (FODs) in Windows 11, navigate to Settings > Apps > Optional features > Add an optional feature. Select View features and search for “RSAT.”
Windows 11 Enterprise Evaluation – Interested in trying Windows 11 Enterprise or Windows 11 Enterprise LTSC on behalf of your organization? Download a free 90-day evaluation.
Windows 11, version 24H2 update history – Quickly access release notes for monthly security updates, optional non-security releases, and out-of-band updates.
Windows release health – Find information about known issues, resolutions, and safeguards to help you plan your updates. Bookmark the Windows message center for updated-related news and announcements. If you are a Windows admin, you can access more details and subscribe to notifications in the Microsoft 365 admin center.

We are working to quickly update the Windows 11 Deployment Lab Kit, which provides an automatically provisioned lab environment and lab guides to help you plan, test, and validate your deployment. Follow us on LinkedIn or @MSWindowsITPro to be informed when the updated kit is available.

The Long-Term Servicing Channel (LTSC)

For organizations with special-purpose devices and environments, such as manufacturing or healthcare systems, we have also released two versions of Windows 11 to the Long-Term Servicing Channel: Windows 11 Enterprise LTSC 2024 and Windows 11 IoT Enterprise LTSC 2024. The LTSC is intended for fixed-function, special-purpose devices that require a longer support lifecycle. Windows 11 Enterprise LTSC 2024 is available as a per user or per device model, depending on the Volume Licensing program through which it is acquired. Windows 11 IoT Enterprise LTSC 2024 is available through original equipment manufacturers (OEMs) or directly from Microsoft through Volume Licensing.

Make updating Windows easier

If you’d like to spend less time managing updates, now is a great time to look at Windows Autopatch. Included with Windows Enterprise E3 (or higher), Windows Autopatch makes it easier to manage feature and monthly security updates for Windows, as well as updates for Microsoft 365 Apps for enterprise, Microsoft Edge, and Microsoft Teams. To help you take a closer look at this cloud service, watch our recent episode of Tackling Tech.

For additional help, tips, and best practices, join Windows Office Hours every third Thursday on the Windows Tech Community! We assemble experts from the Windows, Intune, and Windows 365 product teams, as well as adoption specialists and Fast Track to answer your questions in chat. Join us Thursday, October 17 and bookmark https://aka.ms/Windows/OfficeHours for upcoming dates.

Frequently asked questions

Is there an enablement package for this release?
No. Windows 11, version 24H2 requires a full OS swap so it cannot be deployed using servicing technology. In addition, devices must be running Windows 11, version 23H2 or 22H2 with the May 2024 non-security preview update installed in order to update to version 24H2.

Can I upgrade our organization’s devices from Windows 10 directly to this new update?
Yes. You can upgrade devices directly from Windows 10 to Windows 11, version 24H2[4] using the target version capability in Windows Update for Business, the Windows Update for Business deployment service, and feature update deployments in Windows Autopatch. In fact, we strongly encourage you to begin your upgrade immediately, as Windows 10 will reach end of support on October 14, 2025. For end user readiness, take advantage of our Windows 11 Onboarding Kit, which provides ready-to-customize materials you can use to help prepare your users and ensure that they get the most out of Windows 11. If you need a little more time to get from Windows 10 to Windows 11, or procure new devices, you can also explore the Windows 10 Extended Security Updates (ESU) program.

How can I test new features in Windows 11 prior to their release?
Join the Windows Insider Program. Access preview builds of Windows 11 so you can try out the new features, provide feedback directly to Microsoft, and, if desired, deploy early builds across your organization to test your apps, hardware, and processes. Want to work directly with the Windows engineering teams to develop solutions that better meet your needs? Join the Microsoft Management Customer Connection Program.

What if my applications are not compatible with Windows 11, version 24H2?
Windows 11 is the most compatible Windows operating system ever. If you do encounter an application compatibility issue, however, our App Assure team is here to help.

Is there a Windows Server release with this release of Windows 11?
Windows Server 2025 is currently in preview. To learn more, see What’s new in Windows Server 2025 (preview).

Are there new features for Copilot+ PCs?
Copilot+ PCs run on Windows 11, version 24H2, enabling us to unlock productivity, accessibility, and sustainability capabilities that are exclusive to Copilot+ devices. We introduced Copilot+ PCs with features like Live Translation and Paint Cocreator and announced new features today[5] that will begin rolling out to Windows Insiders in phases, to select devices and markets, beginning in November. Our goal is to ensure that IT admins have the controls they need to confidently deploy Copilot+ PCs and the exclusive AI features that run on them. With this latest update, we have provided additional controls across Copilot+ PC features and have even more coming to ensure these AI experiences can be deployed by IT in organizations. To hear more about what is coming for Copilot+ PCs, and how you as the IT admin will be in control, tune in to Microsoft Ignite this November.

What changes have you made to the security posture of Recall on Copilot+ PCs?
Recall is an entirely new way to instantly find something you’ve previously seen on your PC. Based on your feedback, we have worked to make Recall[6] even more secure by default, ensuring that you feel confident that your data is safe and secure. Additionally, Recall will be off by default for all commercial[7] devices. You will need to allow Recall via policy before users will be able to opt-in. Once enabled, devices will use Windows Hello to confirm a person’s identity (and presence) and then securely unlock the Recall experience. For more details, see the recent Update on Recall security and privacy architecture. We will share more about controls for Recall and other Copilot+ PC news at Microsoft Ignite.

Additional resources

For more details on the rollout strategy for Windows 11, version 24H2 and new features for Copilot+ PCs, see How to get new experiences for Windows 11.
For insight into the update experience for home users, see How to get the Windows 11 2024 Update.
For a closer look at all the latest features and enhancements, see New experiences coming to Copilot+ PCs and Windows 11.

Continue the conversation. Find best practices. Bookmark the Windows Tech Community, then follow us @MSWindowsITPro and on LinkedIn. Looking for support? Visit Windows on Microsoft Q&A.

[1] Downloads in the Microsoft 365 admin center and similar channels may be delayed.

[2] Home and Pro editions of Windows 11, version 24H2 receive 24 months of support. For more details, see the Windows lifecycle FAQ.

[3] Content Adaptive Brightness Control (CABC) is not available on all laptop or 2-in-1 devices as it must be enabled by the manufacturer of the device in order to function.

[4] If we identify an issue that might cause a feature update to fail or roll back, a safeguard hold may be applied to prevent affected devices from installing the update and safeguard them from a poor experience.

[5] Availability may differ based on the various silicon platforms.

[6] Starting in October, Windows Insiders will begin to see Recall on Snapdragon-powered Copilot+ PCs. Those who purchase Intel- or AMD- powered Copilot+ PCs will initially see Recall in the Windows Insider Program starting in November. Timing and details on the broad availability of Recall will be shared at a later date. Security continues to be our top priority and when Recall is available for Windows Insiders, we will publish a blog with more details.

[7] Commercial devices are defined as devices running an Enterprise or Education SKU, devices running a premium SKU that are managed by an IT administrator (whether via Microsoft Intune or other endpoint management solution), devices that have a volume license key, or devices that are joined to a domain.

​Microsoft Tech Community – Latest Blogs –Read More