Opinion

Joe Carlyle

From Ops to C-Level – AI Adoption in Practice – Festive Tech Calendar 2025

Introduction

It’s great to be back for another year of the Festive Tech Calendar. Always an excellent event across the month of December, and looking through the schedule, it is an amazing year of content. Kudos to the team for all the work they put in, and thanks for having me again!

This year, I’m continuing my AI adoption theme, but expanding it slightly beyond the scope of Azure, and taking a bit of a ‘business via tech’ approach. This is for a couple of reasons; first, I think AI has changed in the last 12 months, and second, I think the way everyone approaches it also has. This might seem simple, but in reality, or in practice for an IT decision maker – it really isn’t. So, I thought I would share some learnings from where I have seen things work well and work not so well!

AI Opinion

This of course all starts with AI. If you don’t know what that is, I would be slightly worried as to how you found yourself reading this… but AI isn’t just one thing anymore, and to be honest, it never was to begin with.

The GPT models have become the Google of search engine colloquialism for common users – ‘just ask ChatGPT’ etc. This is great for awareness, but also no doubt frustrating for the teams creating Copilot, Gemini etc. Even more frustrating for the teams responsible for the original Azure AI Services (Cognitive Services – remember that!?). That brings me onto my next point, and one of the key challenges with AI adoption I have seen – AI isn’t just generative and the perception gap that opens up.

AI ‘Types’ I Think are Relevant

For most users, working with an AI assistant of some sort, is the majority of AI interaction they will have. So general and/or company based knowledge helping users find or create content that assists with work. I genuinely haven’t seen any work setting where stupid images etc are being created, so don’t believe the hype or prevalence of that based on social media.

Next, a subset of your user base may use more tuned AI assistance. Think Github Copilot as the example here. Chat based, powerful and specific AI based assistance. This is often seen as a more impactful ‘upgrade’ to a user/teams skillset, but much less clearer in terms of adoption requirements.

Then we move into the one-to-many AI – agents. A huge range of options here and with the capabilities in Azure AI Foundry, a near endless set of use cases. From experience, I’ve seen these used to take on one information source, or one task. These work well, roll out rapidly and require little to no real guidance. I have also seen attempts at multi-agent frameworks/workflows with less success, and finally very few agents that take action without explicit supervision. Take an over simplistic example – “Write a reminder email for the upcoming meeting and send to all attendees, cc my boss” you need serious confidence in that agent to handle information as you need it to.

Finally, there has been large scale adoption of AI within systems or existing applications. However, don’t mistake large scale adoption with actual success of AI. This is easily the example where I have seen the most ‘AI washing’ – reinventing some existing feature by calling it AI. This part really bugs me, as I believe it is driving up costs of already expensive systems, while also interrupting roadmaps of upgrades, fixes, new features that could have more impact.

Ok – let’s get into some points around adoption in practice. Ultimately I’ve seen it boil down to a balance of use case vs blockers. If you can outdo the list of blockers, and the use case is valid – success. Now I have drastically simplified that, but let’s start with blockers, so we can start to see the meat of the issue.

Blockers

When C-level look for AI adoption, they think of productivity return-on-investment, people management, time to market, and competition. This is no different than any other type of adoption, but I think AI carries a much tougher perception problem. In a presentation back in the Summer, and in every AI discussion since, I have discussed and named this as the ‘surely gap’.

Without exception this is the number one issue derailing AI adoption. If you cannot address the perception issue early, you are doomed to have a project that will be viewed as not meeting expectations, disappointing, or even a failure. Even though AI might be delivering 50-80% increases in productivity, output, or accuracy etc. The second you hear “Surely AI can solve this, surely AI can create that, surely AI can remove that team – you are surely, in trouble.

Flip the approach to the ops team, or IT people. I see two equal priority issues:

  • Bad data projects – “I’d love to use AI, but our data isn’t ready” – this is either a security issue, or a compliance issue, or both. Often it can be as simple as permissions and structure of data, commonly an issue in any SharePoint online environment that has been around a while. Plenty of simple work can address this, but the fear of sensitive data being exposed will put the brakes on any AI work. Now a positive swing here is perhaps you can now get that data project budget based on AI, but it’s still a significant challenge.
  • Sprawling use cases – this is causing more of a compliance and regulatory issue, with no real resolution I have seen, only mitigation via policy. Take a system that can do five things, but your users only need two of them. So you risk assess and secure those two. However, if you can’t disable the other three features, users can simply use them if they wish. And it might not be as simple as features, it becomes more complex with generative AI. I expect changes in monitoring, analytics and feature decoupling to come as the EU AI act takes hold.

Lessons Learned

The first challenge with any blocker is knowing about it. Once you know about a problem, you can start to formulate a plan to solve it. And with the blockers I’ve outlined, you can probably already guess some solutions.

First and most important in my experience is dealing with the perception issue. I view AI as the least IT project a company can take on at present. Something like Copilot has little to no real IT or Ops work to configure it. License the user – milestone complete. But if an Ops team is given this as a project it can miss the real beginning an end of the project – perception and adoption.

Address the perception up front – why do we need AI, what will it do, what does good look like? Work backwards, and pick use cases that have simple, measurable outcomes.

Plan the adoption the second you have defined use cases – targeted users, timeframe, and a cadence of revisit. Most AI usage requires habit forming of some sort, adoption plans need to promote and push that.

In terms of Ops challenges, the most important lesson I have learned is get the team onboard, and get them out of the way. AI has the worst risk of shadow IT and data leakage I have ever seen. Users will want to try and use it. Give them something early, give them something decent and give them guidance. Then and only then – block everything else you can.

My Wishlist

This is tricky, but maybe not too complex. Instead of a list, I think I would push for one change and one change only – greater awareness or understanding of how to align a use case and capability. Nobody thinks email is the entire answer to their comms requirement, and they aren’t disappointed when it can’t do instant messaging or video calling. I know that isn’t necessarily a fair comparison, but if we can get AI understanding closer to that, I think it will greatly improve not only the adoption rates, but the success of AI projects.

I have another point on cost, and probably several points on environmental impact, but they are at least a blog post each. Perhaps something on that in Azure Spring Clean 2026…

In Closing

To close I will convert my opinion to a simple list:

  • Speak to your stakeholders, gather opinions and use cases. Identify enablers and detractors, plan for both.
  • Pick your use cases, always start with the simple ones that have measurable confirmed outcomes.
  • Address Ops concerns, get the team onboard for rollout. Create your plan for enablement and adoption.
  • Meet your stakeholders again, get the use case and outcome crystal clear. Leave no room for the ‘surely gap’
  • Rollout and continuously adopt. Revisit use and the use case. Evolve the plan as the use case does.
Joe Carlyle

Microsoft MVP Summit 2024

This week was the MVP Summit for 2024. As with all previous events, all content is completely under NDA. That means this post isn’t going to disclose any of the information that was discussed, but I can mention some areas/ideas that I am excited about.

Most years, Microsoft runs several conferences. Some are more impactful than others, and this year is no different. I am really looking forward to Build and the developer focussed content that will accompany it. And not since 2019 am I looking forward to Ignite as much as I am this year. It appears to be almost back to its full strength, being run this year in Chicago across several days.

While we were at Summit, the GA of Copilot for Security was announced for April 1st, another major milestone for Microsoft’s Copilot products. It doesn’t take much foresight to guess that this pace and innovation of AI products will continue into 2024.

One non-technical item that was very important was my fellow MVPs. This year had an increased in-person turnout to last year, and everyone was in fantastic form for the week. As evidenced in the photo below which includes myself, and other Irish MVPs we could round up on short notice for a quick picture at Microsoft’s new building courtyard.

And for the first time ever, I was joined by a colleague! Alan Kinane is new into our team, and it was great to be able to chat on different levels about content.

And lastly, a massive thank you to those who organised, ran, and made sure everything went as planned for the week. I can only imagine how complex it all is. It’s very much appreciated!

Joe Carlyle

What are – Microsoft Applied Skills

Last month, Microsoft introduced a new method of verifying your capabilities when it comes to Microsoft technology, Applied Skills. Critically, Applied Skills is focused on verifying hands-on experience. See the blog post announcement here.

Not long after the change from MCSA/MCSE to Role Based Certifications a few years ago, a section of questions based on a lab environment was introduced. This didn’t last very long, and had several teething issues. However, I was a fan of the attempt. As a result, I am delighted to see something similar being reintroduced. Funnily enough, I also like the fact that you can gain an Applied Skill credential from home, open book. We all work in an open book where Google/Bing etc are our sidekicks in sanity checking an error, ensuring that setting/parameter is as you remember it or looking up something new. It doesn’t take away from the experience needed to work with the technology.

I also like that Microsoft Learn are presenting this as a parallel, somewhat complimentary channel to Certifications. And of course, I love that they are online verifiable so that they can form part of your CV/resume. And to be honest, as someone who works with many technical peers on my team, while I know that credentials like this do not guarantee someone is good at the job or has the exact correct experience, I am at the point where if someone is good at their job and does have the skills, it is more odd to me that they haven’t simply passed all the relevant exams – it’s easy, no?

At launch there were several Applied Skills to achieve, at Ignite last week, several more were added, and there are more to come. The below is the current poster advertising what’s possible across pillars

Let’s start with some simple advice – When I first saw this launch, I was excited and clicked through to the secure networking skill (#AzNet all the way people) using my phone, while sitting on my couch. This loaded the assessment window and launched the lab – of which I could see nothing. The screen is far too small to function, and I really wasn’t paying proper attention. However, even without doing anything and simply exiting, it counted as an attempt, and I couldn’t retry for 72 hours. Don’t repeat my mistake, use a computer!

Ok, the assessment/lab itself – I liked it. In fact, I don’t think I could fault it. It loaded quickly, instructions are clear, results are immediate. My only gripe to date is that the results aren’t detailed enough. I was a few points shy of perfect for the secure networking skill (when I sat it properly 🙂 ) and the results are all green ticks, so I have no idea which element was incorrect, or if I missed something. Once loaded, you have a full two hours to complete the assessment, which may seem like a lot, but not if you’re not prepared. I’ve sat several of these now, and what you are being asked to do ranges from simple configuration tick boxes, to complex, layered implementation. The complex tasks ask for a simple result, you need to know how to get there. Without experience, you will struggle to figure this out via Google within your time window, so do the prep work! I found this out personally when I had to figure out some Python for the Document Intelligence assessment, but thankfully I still passed.

As someone who sees great value in having these available, free, to everyone, I think this is an excellent addition to Microsoft Learn. I’ve sat and passed four so far, and intend to continue with the areas I already know and expand into those that I don’t. I also intend to continue to sit new exams and renew all of my Certifications as well. One thing with Microsoft, and specifically Azure – never stop learning!

Joe Carlyle

Can Azure OpenAI Write This Blog Post?

This is the fear right? Or maybe it’s not a fear, and is a desire for some people, but it’s still not clear what the impact may be. What am I talking about? AI of course; specifically for this post, Azure OpenAI. Now, my idea here is to see what I think of the quality of responses it can give me based on specific prompts to try and tie together a simple blog post. I will pick a topic, an objective for the post, and I will be 100% transparent on what is AI written and what is me – just in case it is not obvious! I will also include the settings being used for my model at the end and whether I adjusted them during.

Ok, let’s get started. My topic is going to be Virtual Network Manager in my usual style of exploration.

What is Azure Virtual Network Manager? Write an opening paragraph for a blog post.

Azure Virtual Network Manager is an essential tool for managing and organizing your cloud-based virtual networks. With virtual networks, you can securely connect and isolate your Azure resources, control traffic flow, and expand your on-premises network into the cloud. However, managing virtual networks can be a challenging task, especially when you have multiple resources and services scattered across different regions. This is where the Azure Virtual Network Manager comes in, offering a centralized and user-friendly interface to help you efficiently manage and monitor your virtual networks. In this blog post, we will take a closer look at the Azure Virtual Network Manager and explore its features and benefits.

What are the top three use cases for Azure Virtual Network Manager? Write a detailed paragraph including bullet points and examples for each.

Azure Virtual Network Manager is a versatile tool that can support various use cases, depending on your organization’s needs. Here are the top three use cases for Azure Virtual Network Manager:

  1. Hybrid Cloud Connectivity: Azure Virtual Network Manager can help organizations establish secure, high-speed connections between their on-premises datacenters and Azure resources. This allows companies to take advantage of the scalability and flexibility of the cloud while maintaining the security and control of their own infrastructure. With Azure Virtual Network Manager, organizations can easily configure VPN gateways, express routes, and other connectivity options to seamlessly integrate their on-premises networks with Azure.

    Examples:
  • Connecting an on-premises Active Directory to Azure to manage identity and access for cloud resources.
  • Extending a corporate LAN to Azure to host applications and services in the cloud.
  • Using Azure as a backup and disaster recovery site for on-premises datacenters.
  1. Multi-Tier Application Architecture: Azure Virtual Network Manager enables organizations to create complex multi-tier application architectures in the cloud, with the ability to isolate and secure different layers of the application. This is done by creating multiple virtual networks, each with its own subnets and security groups. This approach provides better control over network traffic and reduces the risk of security breaches by limiting the attack surface.

    Examples:
  • Hosting a web application in Azure with a separate virtual network for the web server, application server, and database server.
  • Creating a DMZ in Azure to host public-facing services, with a separate virtual network for internal services.
  • Separating development, testing, and production environments into different virtual networks to ensure isolation and security.
  1. Global Network Management: Azure Virtual Network Manager provides a centralized interface to manage virtual networks across different regions and subscriptions. This allows organizations to monitor network traffic, troubleshoot connectivity issues, and optimize network performance from a single location. With Azure Virtual Network Manager, organizations can also implement advanced networking features, such as traffic routing, load balancing, and network security groups.

    Examples:
  • Creating a global load-balancing solution to distribute traffic across multiple regions and availability zones.
  • Implementing Azure Firewall to provide centralized network security management across multiple virtual networks.
  • Enabling Azure Traffic Manager to route traffic to the nearest endpoint based on geographic location or network latency.

What are the current limitations or known issues of Azure Virtual Network Manager? Write a paragraph in the blog post.

While Azure Virtual Network Manager is a powerful tool for managing virtual networks in the cloud, there are some limitations and known issues that organizations should be aware of. Some of the current limitations of Azure Virtual Network Manager include:

  • Limited support for hybrid networking configurations: Currently, Azure Virtual Network Manager has limited support for hybrid networking configurations, which can make it challenging to manage virtual networks that span multiple cloud providers or on-premises datacenters.
  • Limited visibility into network traffic: Azure Virtual Network Manager provides limited visibility into network traffic, which can make it difficult to troubleshoot connectivity issues or monitor network performance.
  • Limited integration with third-party tools: Azure Virtual Network Manager has limited integration with third-party tools, which can make it challenging to implement advanced networking features or security solutions.

    Organizations should carefully evaluate their networking needs and consider these limitations before deploying Azure Virtual Network Manager. However, Microsoft is continually improving the tool and addressing these limitations through updates and new features.

Is there a tutorial or general steps that should be recommended as part of this blog post for people who are interested in now trying Azure Virtual Network Manager? Include high level steps and a link to the appropriate site if possible

Yes, Microsoft offers a comprehensive tutorial on how to set up and manage virtual networks using Azure Virtual Network Manager. Here are the high-level steps for getting started with Azure Virtual Network Manager:

  1. Create a virtual network: The first step is to create a virtual network in Azure. This involves specifying the name, address space, and subnets for your network. You can also configure advanced network settings, such as DNS servers and network security groups.
  2. Create a gateway subnet: If you plan to connect your virtual network to an on-premises network, you need to create a gateway subnet. This subnet is used to deploy a VPN gateway or an ExpressRoute gateway.
  3. Configure VPN or ExpressRoute: Once you have created your virtual network and gateway subnet, you can configure a VPN or ExpressRoute connection to connect your on-premises network to Azure.
  4. Set up network security: Azure Virtual Network Manager provides several tools for securing your virtual network, including network security groups, application security groups, and Azure Firewall. You can use these tools to control inbound and outbound traffic, restrict access to specific resources, and monitor network activity.
  5. Monitor and troubleshoot your network: Azure Virtual Network Manager provides several tools for monitoring and troubleshooting your virtual network, including network performance monitoring, packet capture, and network watcher.

    For more detailed instructions on how to use Azure Virtual Network Manager, check out the official Microsoft tutorial: https://docs.microsoft.com/en-us/azure/virtual-network/tutorial-create-virtual-network.

Based on your previous responses do you think there is anything else that is important to know about Virtual Network Manager? Write a closing paragraph with a call to action to get in touch if you have any questions

In conclusion, Azure Virtual Network Manager is a powerful tool that can help organizations manage and secure their virtual networks in the cloud. With Azure Virtual Network Manager, organizations can establish secure connections between their on-premises and cloud-based resources, create complex multi-tier application architectures, and manage virtual networks across different regions and subscriptions. While there are some limitations and known issues with the tool, Microsoft is continually improving it with updates and new features.

If you have any questions about how to use Azure Virtual Network Manager or want to learn more about how it can support your organization’s networking needs, don’t hesitate to get in touch. Our team of experts is always available to help you navigate the complexities of cloud networking and find the best solutions for your business.

Let’s start with the settings used, I started with these and didn’t change them at all during the process:

engine="35t", (gpt35-turbo)
temperature=0.8,
max_tokens=800,
top_p=0.95,
frequency_penalty=0,
presence_penalty=0

Overall, I think it has the basics nearly correct. However, it has very little personalisation, it obviously speak to no real experience, and is essentially a modified read of the Docs pages. So I am pretty confident that at no point in near future could it be used to write an entire post that I would be happy with publishing. However, where a post calls for something more generic, like known issues, or intros, it could definitely be used as a starting point.

I am interest to see if a GPT4 model would make a huge difference, but for this style of work, I am not sure. Also, I could of course edit the temperature or top p, but I am unsure as to whether that impact again would be noticeable enough.

In conclusion, Azure OpenAI can definitely write a blog post – I am just not sure it’s a blog post I would be happy with!