Hi,

I`m planning to buy System Center and have some doubts about witch licences i need.

I have a 3 node vmware cluster with 96 total physycal cores.

I’m planning to use configuration manager to manage deployment, updates and app installation on near 400 client pc.

As i understand if i buy datacenter edition i can manage all windows servers on the cluster and have to buy 400 operations manager client licences for client pc’s and with standard i have to buy client licences for servers to.

Is that ok? 

Thanks

​Hi,I`m planning to buy System Center and have some doubts about witch licences i need.I have a 3 node vmware cluster with 96 total physycal cores.I’m planning to use configuration manager to manage deployment, updates and app installation on near 400 client pc. As i understand if i buy datacenter edition i can manage all windows servers on the cluster and have to buy 400 operations manager client licences for client pc’s and with standard i have to buy client licences for servers to.Is that ok?  Thanks  Read More

​

I have a 40 slide presentation. Some of the slides are particularly wordy, so what I was planning to do was to ask Copilot to summarise a series of slides and add that as a companion document to the presentation.

When I ask it to summarise a single slide by asking it to “Summarise slide 5” it grabs a title from a previous headline slide and creates a summary based seemingly on the whole presentation. If I ask it to summarise multiple slides with “Please summarise slides 5, 7, 10….” It creates a headline for each slide but summarises the wrong slides throughout and places the wrong titles on each slide in the response.

Anyone else tried summarising specific slides in a presentation?

​I have a 40 slide presentation. Some of the slides are particularly wordy, so what I was planning to do was to ask Copilot to summarise a series of slides and add that as a companion document to the presentation. When I ask it to summarise a single slide by asking it to “Summarise slide 5” it grabs a title from a previous headline slide and creates a summary based seemingly on the whole presentation. If I ask it to summarise multiple slides with “Please summarise slides 5, 7, 10….” It creates a headline for each slide but summarises the wrong slides throughout and places the wrong titles on each slide in the response. Anyone else tried summarising specific slides in a presentation?  Read More

​

Hi all

I am wanting to put a script within Intune so during the deployment phase of autopilot it would remove the HP Bloatware however if I attach this script

https://gist.github.com/mark05e/a79221b4245962a477a49eb281d97388 into the deployment policy it would fail everytime.

I can only remove it by going to the start menu , run powershell as admin and then run the script manually from there.

What do I need to add so the script can run automatically?

It wouldn’t matter if it can’t be run during the enrollment stage but I do want to somehow automatic this .

Thanks

​Hi all I am wanting to put a script within Intune so during the deployment phase of autopilot it would remove the HP Bloatware however if I attach this scripthttps://gist.github.com/mark05e/a79221b4245962a477a49eb281d97388 into the deployment policy it would fail everytime. I can only remove it by going to the start menu , run powershell as admin and then run the script manually from there. What do I need to add so the script can run automatically? It wouldn’t matter if it can’t be run during the enrollment stage but I do want to somehow automatic this . Thanks  Read More

​

Overview

Amazon Web Services (AWS) has announced the discontinuation of its Amazon Quantum Ledger Database (QLDB). Their documentation has been updated to indicate that support for QLDB will end on July 31, 2025. AWS’s decision has prompted many customers to explore alternative solutions for their ledger database needs.

Introducing ledger in Azure SQL

AWS propose Aurora PostgreSQL for audit use cases as an alternative to QLDB for common ledger database use cases. While Aurora offers detailed audit logging and permanent log retention, it does not include cryptographic verifiability.

Microsoft offers an excellent replacement through its ledger feature in Azure SQL. It provides similar functionalities, ensuring data integrity through cryptographic verification, and leverages the full capabilities of Azure SQL. Ledger is a feature that offers the power of Blockchain in Azure SQL Database, Azure SQL Managed Instance and SQL Server. Ledger allows establishing trust across different business entities while maintaining the simplicity and performance of a relational database. The data is centrally managed, and you can cryptographically attest to other parties, such as auditors or business partners, that your data can be trusted and hasn’t been tampered with.

More information about ledger in Azure SQL and SQL Server can be found in the ledger documentation.

Key features of ledger

Captures and cryptographically links data changes to make the data tamper-evident and verifiable.
Uses the same SQL environment you already know.
Is easy to deploy and maintain.
It comes with NO extra cost.

Migration support

Amazon QLDB uses a document-oriented data model. The complexity lies in normalizing data during migration, converting the document model into a relational model, and handling any changes to the document model. Microsoft can assist organizations in migrating from Amazon QLDB to Azure SQL Database. For more detailed information and to initiate the migration process, organizations are encouraged to contact Microsoft support or their Microsoft Account Manager.

Conclusion

Even though QLDB is being phased out, Microsoft offers an excellent alternative for users to host their data with cryptographic immutability. This ensures they can uphold rigorous data integrity standards. Ledger in Azure SQL serves as a strong alternative, with effortless integration into the Azure SQL environment.

Useful links

For more information and to get started with ledger in Azure SQL, see:

Explore the Azure SQL Database ledger documentation
Read the whitepaper
GitHub demo/sample
Data Exposed episode (video)
Listen to the ledger podcast

​Microsoft Tech Community – Latest Blogs –Read More 

Hi,

Since the new teams i get sometimes the exclamation mark and i have an “account problem” with one of my customers. Normally my signin is expired there and i need to resign in.

But i do not need them at the moment. The exclamation mark kills the “unread messages” number on my taskbar – which is needed. Anyone else experiencing this? Signing in to the customer tenant(s) is no real solution – as this disrupts the work flow.

BR

Stephan

​Hi, Since the new teams i get sometimes the exclamation mark and i have an “account problem” with one of my customers. Normally my signin is expired there and i need to resign in. But i do not need them at the moment. The exclamation mark kills the “unread messages” number on my taskbar – which is needed. Anyone else experiencing this? Signing in to the customer tenant(s) is no real solution – as this disrupts the work flow. BRStephan  Read More

​

Hello everyone,
Yesterday I downloaded the iso of the latest 24h2 release preview and installed it, then chose the beta channel as the insider subscription type, but when I restarted it didn’t keep the selection because I saw myself in release preview without the possibility of choosing the beta channel, which in fact is grayed out and not selectable, while dev and canary can be chosen.
Currently, I am in release preview with build 26100.1742 ge_release.
I installed all the updates and restarted again, but nothing has changed.
Can you tell me why and where I went wrong?
Thanks everyone.

​Hello everyone,Yesterday I downloaded the iso of the latest 24h2 release preview and installed it, then chose the beta channel as the insider subscription type, but when I restarted it didn’t keep the selection because I saw myself in release preview without the possibility of choosing the beta channel, which in fact is grayed out and not selectable, while dev and canary can be chosen.Currently, I am in release preview with build 26100.1742 ge_release.I installed all the updates and restarted again, but nothing has changed.Can you tell me why and where I went wrong?Thanks everyone.  Read More

​

Hi everyone,

I’m currently working on a project in Azure Synapse where I’m using the SAP CDC Connector to connect to an S4Hana system. My goal is to filter data on the source side before storing it in my ADLS Gen2, as there are certain data restrictions that I need to adhere to.

I need to fetch multiple objects from SAP, and I typically use a parameterized approach for this. I have a JSON file that contains parameters and queries for each object I want to retrieve from the source. For instance, I define SQL queries in the JSON file to perform the filtering. This method works well with SQL Connectors.

However, with the SAP CDC Connector, I haven’t been able to find any functionality that allows me to apply such filtering directly at the source.

Here’s what I’m doing so far:

I’m currently using a dataflow in a for each loop. In the dataflow however, I cannot pass SQL queries and Im stuck with the expression builder. I cannot figure out how to dynamically pass query like filtering. So Im just getting the unfiltered objects, which is not an option. I have so many objects, that I cant maintain a non parameterized version.

I tried using a copy data activity as well, however when selecting it, I do not get the option to choose the SAP CDC Integration Dataset.

Has anyone successfully managed to filter tables at the source when using the SAP CDC linked service? Any insights or suggestions on how to achieve this would be greatly appreciated.

Thanks in advance for your help!

​Hi everyone, I’m currently working on a project in Azure Synapse where I’m using the SAP CDC Connector to connect to an S4Hana system. My goal is to filter data on the source side before storing it in my ADLS Gen2, as there are certain data restrictions that I need to adhere to.I need to fetch multiple objects from SAP, and I typically use a parameterized approach for this. I have a JSON file that contains parameters and queries for each object I want to retrieve from the source. For instance, I define SQL queries in the JSON file to perform the filtering. This method works well with SQL Connectors.However, with the SAP CDC Connector, I haven’t been able to find any functionality that allows me to apply such filtering directly at the source. Here’s what I’m doing so far:I’m currently using a dataflow in a for each loop. In the dataflow however, I cannot pass SQL queries and Im stuck with the expression builder. I cannot figure out how to dynamically pass query like filtering. So Im just getting the unfiltered objects, which is not an option. I have so many objects, that I cant maintain a non parameterized version.I tried using a copy data activity as well, however when selecting it, I do not get the option to choose the SAP CDC Integration Dataset. Has anyone successfully managed to filter tables at the source when using the SAP CDC linked service? Any insights or suggestions on how to achieve this would be greatly appreciated. Thanks in advance for your help!  Read More

​

Hi

I had Entra Connect running for a long time without issues. Out of the blue Connect Sync started to report Duplicate Attribute on 3 users User Principal Name.

The 3 users, Connect Sync believe has a conflicting value in Entra, do exist in Entra, but with a smtp address which matches the UPN, and is not the the users UPN. 

If i run the following command on my on-prem AD the UPN does not exist in any form of domain name:

Get-ADUser -Filter {UserPrincipalName -eq “email address removed for privacy reasons”}

Get-ADUser -Filter {UserPrincipalName -eq “e-mail@domain.local”}

Get-ADUser -Filter {UserPrincipalName -eq “email address removed for privacy reasons”}

All my users UPN are different from the configured on-prem ProxyAddresses so the above error mesage makes no sense. And futher more the 3 users which sync sees as a conflict do not even has a ProxyAddresses configured.

Any ideas how to futher debug this?

/Robert

​Hi I had Entra Connect running for a long time without issues. Out of the blue Connect Sync started to report Duplicate Attribute on 3 users User Principal Name. The 3 users, Connect Sync believe has a conflicting value in Entra, do exist in Entra, but with a smtp address which matches the UPN, and is not the the users UPN.  If i run the following command on my on-prem AD the UPN does not exist in any form of domain name: Get-ADUser -Filter {UserPrincipalName -eq “email address removed for privacy reasons”}Get-ADUser -Filter {UserPrincipalName -eq “e-mail@domain.local”}Get-ADUser -Filter {UserPrincipalName -eq “email address removed for privacy reasons”} All my users UPN are different from the configured on-prem ProxyAddresses so the above error mesage makes no sense. And futher more the 3 users which sync sees as a conflict do not even has a ProxyAddresses configured. Any ideas how to futher debug this? /Robert        Read More

​

boa noite, hoje eu fui usar a função histórico de ações no Excel do meu pc, mais sempre da bloqueado, e quando eu vou ver aparece que preciso de licença, vocês sabem o que isso significa? espero que possam me ajudar.

​boa noite, hoje eu fui usar a função histórico de ações no Excel do meu pc, mais sempre da bloqueado, e quando eu vou ver aparece que preciso de licença, vocês sabem o que isso significa? espero que possam me ajudar.  Read More

​

The rapid evolution of AI has led to powerful tools for knowledge retrieval and question-answering systems, particularly with the rise of Retrieval-Augmented Generation (RAG) systems. This blog post introduces my capstone project, created as part of the IXN program at UCL in collaboration with Microsoft, aimed at enhancing RAG systems by integrating multimodal knowledge extraction and retrieval capabilities. The system enables AI agents to process both textual and visual data, offering more accurate and contextually relevant responses. In this post, I’ll walk you through the project’s goals, development journey, technical implementation, and outcomes.

Project Overview

The main goal of this project was to improve the performance of RAG systems by refining how multimodal data is extracted, stored, and retrieved. Current RAG systems primarily rely on text-based data, which limits their ability to generate accurate responses when queries require a combination of text and images. To address this, I developed a system capable of extracting, processing, and retrieving multimodal data from Wikimedia, allowing AI agents to generate more accurate, grounded and contextually relevant answers.

Key features include:

Multimodal Knowledge Extraction: Data from Wikimedia (text, images, tables) is preprocessed, run through the transformation pipeline, and stored in vector and graph databases for efficient retrieval.

Dynamic Knowledge Retrieval: A custom query engine, combined with an agentic approach using the ReAct agent, ensures flexible and accurate retrieval of information by dynamically selecting the best tools and strategies for each query.

The project began by addressing the limitations of existing RAG systems, particularly their difficulties with handling visual data and delivering accurate responses. After reviewing various technologies, a system architecture was developed to support both text and image data. Throughout the process, components were refined to ensure compatibility between LlamaIndex, Qdrant, and Neo4j, while optimising performance for managing large datasets. The primary challenges lay in handling the large volumes of data from Wikimedia, especially the processing of images, and refactoring the system for Dockerisation. These challenges were met through iterative improvements to the system architecture, ensuring efficient multimodal data handling and reliable deployment across environments.

Implementation Overview

This project integrates both textual and visual data to enhance RAG systems’ retrieval and response generation. The system’s architecture is split into two main processes:

Knowledge Extraction: Data is fetched from Wikimedia and transformed into embeddings for text and images. These embeddings are stored in Qdrant for efficient retrieval, while Neo4j captures the relationships between the nodes, ensuring the preservation of data structure.

Knowledge Retrieval: A dynamic query engine processes user queries, retrieving data from both Qdrant (using vector search) and Neo4j (via graph traversal). Advanced techniques like query expansion, reranking, and cross-referencing ensure the most relevant information is returned.

System Architecture Diagram

Tech Stack

The following technologies were used to build and deploy the system:

Python: Core programming language for data pipelines

LlamaIndex: Framework for indexing, transforming, and retrieving multimodal data

Qdrant: Vector database for similarity searches based on embeddings

Neo4j: Graph database used to store and manage relationships between data entities

Azure OpenAI (GPT-4O): Used for handling multimodal inputs, deploying models via Azure App Services

Text Embedding Ada-002: Model for generating text embeddings

Azure Computer Vision: Used for generating image embeddings

Gradio: Provides an interactive interface for querying the system

Docker and Docker Compose: Used for containerization and orchestration, ensuring consistent deployment

Implementation Details

Multimodal Knowledge Extraction

The system starts by fetching both textual and visual data from Wikimedia, using the Wikimedia API and web scraping techniques. Then the key steps in knowledge extraction implementation are:

Data Preprocessing: Text is cleaned, images are classified into categories such as plots or images for appropriate handling during later transformations, and tables are structured for easier processing.

Node Creation and Transformation: Initial LlamaIndex nodes are created from this data, which then undergo several transformations through the transformation pipeline using GPT-4O model deployed via Azure OpenAI:

Text and Table Transformations: Text data is cleaned, split into smaller chunks using semantic chunking, and new derived nodes are created from various transformations, like key entity extraction or table analysis. Each node has a unique Llamaindex ID and carries metadata such as title, context, and relationships reflecting the hierarchical structure of the Wikimedia page and parent-child relationships with new transformed nodes.

Image Transformations: Images are processed to generate descriptions, perform plot analysis, and identify key objects based on the image type, resulting in the creation of new text nodes.

Embeddings Generation: The last stage of the pipeline is to generate embeddings for images and transformed text nodes:

Text Embeddings: Generated using the text-embedding-ada-002 model deployed with Azure OpenAI on Azure App Services.

Image Embeddings: Generated using the Azure Computer Vision service.

Storage: Both text and image embeddings are stored in Qdrant with reference node IDs in the payload for fast retrieval. The full nodes and their relationships are stored in Neo4j:

Neo4j graphs (left) and close-up section of the graph (right)

Knowledge Retrieval

The retrieval process involves several key steps:

Query Expansion: The system generates multiple variations of the original query, expanding the search space to capture relevant data.

Vector Search: The expanded queries are passed to Qdrant for a similarity-based search using cosine similarity.

Reranking and Cross-Retrieval: Results are then reranked by relevance. Retrieved nodes from Qdrant contain LlamaIndex node IDs in the payload. These are used to fetch the nodes from Neo4j and then to get the nodes with original data from Wikimedia by traversing the graph, ensuring the final response is based only on original Wikipedia content.

ReAct Agent Integration: The ReAct agent dynamically manages the retrieval process by selecting tools based on the query context. It integrates with the custom-built query engine to balance AI-generated insights with the original data from Neo4j and Qdrant.

Dockerization with Docker Compose

To ensure consistent deployment across different environments, the entire application is containerised using Docker. Docker Compose orchestrates multiple containers, including the knowledge extractor, retriever, Neo4j, and Qdrant services. This setup simplifies the deployment process and enhances scalability.

Docker Containers

Results and Outcomes

The system effectively enhances the grounding and accuracy of responses generated by RAG systems. By incorporating multimodal data, it delivers contextually relevant answers, particularly in scenarios where visual information was critical. The integration of Qdrant and Neo4j proved to be highly efficient, enabling fast retrieval and accurate results.

Additionally, a user-friendly interface built with Gradio allows users to interact with the system and compare the AI-generated responses with standard LLM output, offering an easy way to evaluate the improvements.

Here is a snapshot of the Gradio UI: 

Future Development

Several directions for future development have been identified to further enhance the system’s capabilities:

Agentic Framework Expansion: A future version of the system could incorporate an autonomous tool capable of determining whether the existing knowledge base is sufficient for a query. If the knowledge base is found lacking, the system could automatically initiate a knowledge extraction process to address the gap. This enhancement would bring greater adaptability and self-sufficiency to the system.

Knowledge Graph with Entities: Expanding the knowledge graph to include key entities such as individuals, locations, and events or others appropriate for the domain. This would add considerable depth and precision to the retrieval process. The integration of such entities would provide a more comprehensive and interconnected knowledge base, improving both the relevance and accuracy of results.

Enhanced Multimodality: Future iterations could expand the system’s capabilities in handling image data. This may include adding support for image comparison, object detection, or breaking images down into distinct components. Such features would enable more sophisticated queries and increase the system’s versatility in handling diverse data formats.

Incorporating these advancements will position the system to play an important role in the evolving field of multimodal AI, further bridging the gap between text and visual data integration in knowledge retrieval.

Summary

This project demonstrates the potential of enhancing RAG systems by integrating multimodal data, allowing AI to process both text and images more effectively. Through the use of technologies like LlamaIndex, Qdrant, and Neo4j, the system delivers more grounded, contextually relevant answers at high speed. With a focus on accurate knowledge retrieval and dynamic query handling, the project showcases a significant advancement in AI-driven question-answering systems. For more insights and to explore the project, please visit the GitHub repository.

If you’d like to connect, feel free to reach out to me on LinkedIn.

​Microsoft Tech Community – Latest Blogs –Read More