Michaela Murray

22 July 2020

The IT Architect Interviews #2: The Solution Architect

In the second interview in this series exploring the role of IT Architects, Redgate’s Michaela Murray is joined by Chris Slee, an independent Solution Architect currently working with C5, an IT consultancy in Australia, on a number of projects for the Bank of Queensland. What are the challenges that he comes across when planning major new architectures?

Firstly, Chris, could you tell us what your role is as a Solution Architect?

I move from gig to gig, trying to make a difference wherever I end up. The current project for the Bank of Queensland is to help move their data centers from Brisbane to Sydney. As the Solution Architect, I look after a range of business applications and identify all of the integrations involved and develop the strategies for moving those applications to the new data centers.

How did you become a Solution Architect?

I started my IT career on a helpdesk for an internet service provider and then moved into server support before finding a home in the UNIX and enterprise storage world. I ended up being responsible for the entire data storage environment for the lottery company Tatts Group and then moved on to a Solution Architect role five or so years ago.

I’d reached as far up the ladder as I could on the engineer side and my options were shifting into a management position or continuing to solve problems which is what I like to do. Becoming a Solution Architect was the obvious choice in hindsight, although I can’t say it was obvious at the time.

What do you think it means to be a Solution Architect?

Problem solving is definitely a big part of it now, especially with more and more people wanting off-the-shelf products that don’t need to be heavily customized. It’s about figuring out how to integrate products and I find increasingly that I see this job as being a translator.

I’ll talk to the business side first and understand what they’re trying to do, because IT ultimately is a business tool that exists to achieve a business objective. I like to discover what their pain points are or what’s currently limiting them. Then I take that information and work out the objectives, benefits, limitations, constraints, advantages, assets and liabilities of the project. I put all of that together with a whole bunch of clouds and arrows and boxes with labels on them to translate the business view of what needs to be achieved into a technical view of how it can be delivered. After that, I talk to the technical team about how to make it happen.

One of the areas of tension is that there are so many technologies and products out there and so many ways of solving problems. So one of the things I say is that I don’t have to be the smartest person in the room but I do have to know the people who are. If I think we’re going to solve a business problem with a particular technology, I have to be able to talk to the vendor and be sure it is the right technology. The real skill of the Solution Architect role is being able to talk effectively with both the business side and the technology side.

That leads perfectly to my next question. We’ve found from these interviews that Architects are often described as the bridge between the business and the tech teams. So I wanted to know whether you had any tips for how to become a good bridge between the two when they often talk a very different language.

That’s where the idea of the translator comes in, because you need to talk the language of the audience you have in front of you at that moment. If you can’t speak business language, you’re not going to understand their concerns, and I think that’s the main thing. A business audience will talk about things like ROI, cost reductions, benefit maximization, flexibility and scalability. A technology audience like to talk about the ease of administration, component reuse, failover and recovery, load balances, network zones, firewalls, security policy, and data placement. You need to be able to speak both languages so that you can ask pertinent questions. For example, when you say cost reduction, do you mean you want fewer staff or lower operational expenses?

The problem-solving comes in each conversation where you’re drilling down to specify what they’re actually talking about, because people often use the same terms but they have different meanings. You have to work out where the gaps in the conversation are that need to be filled in so you can get a complete picture of what’s going on, on either the business side or the technology side.

How do you see the role evolving in the next five or ten years? What do you think is going to change?

I suspect the real trick in the future is going to be less around the technical aspects and more about the integration. If you can just dial up SaaS platforms like Azure or Amazon or Google Cloud, you have to work out how to integrate them into a coherent whole that allows you to fulfil a business process. If a customer clicks something on a web page or a mobile device, how do you turn that into an order, a transaction, a product or service fulfilment and then get feedback so you can shape your business to whatever the market is doing?

The other big thing moving forward is that people are starting to realize how valuable their data is, so it’s working out how to get data into an analytics environment. There’s a big push happening at the moment and that’s only going to get bigger. Perhaps more importantly, it’s going to be about the regulations around data, that certain data must be onshore, certain data can be anywhere, certain data has business value or has compliance value, certain data is just transitory and 30 seconds after the data was generated or used it’s no longer relevant and no one cares.

You talk a lot about integration and off-the-shelf products. Do you think we’re moving to a point where software development will be more about using APIs and working out how to integrate things together?

That’s definitely something I’m seeing. I think we’ve all got examples of a core business application that was built 20 years ago and there’s one person in the office who knows how it works and how it’s put together. Everyone wants to get away from that and decrease the time to market, and there’s a vast range of pre-existing APIs out there that will solve all kinds of problems. For instance, if you want to calculate delivery costs to get your product from your distribution center to a customer, most big couriers have an existing API. The world is now so connected that anyone who has a data source can become a data retailer, either with free APIs or subscription APIs.

You mentioned things moving more towards the cloud. When is the right time for people to be moving to the cloud, what should they be considering, and how should it be approached?

When should people be moving to the cloud? The fashionable answer is yesterday. More realistically, I think the mistake people make in moving to the cloud is thinking it’s about simply recreating their data center servers in the cloud. That’s not the way to do it at all.

The first thing people need to do is understand the workloads they’ve got because the cloud is really good for a couple of things – moving the workload closer to your customer base, and scaling workloads up and down.

Secondly, people need to understand whether the workload is actually capable of being moved to the cloud, because there are a lot of legacy apps that can’t do it. They’re built to sit on a piece of tin that has a certain spec and that’s all they can do.

What kind of indicators or measures would you be looking for in a successful cloud migration?

I’m not sure I can give a standard list of metrics because ultimately any move to the cloud should be about solving a business problem.

For instance, if you have an online transaction system that has one big event per year like a Black Friday sale, your physical data center will have been designed to handle that peak workload. For the rest of the year, you’ll have say 80% of your capacity capability sitting idle doing nothing. So moving to the cloud might be about cost reduction. You can run the rest of the year at a reduced rate and scale up to handle the workload during that peak period and then reduce it when the peak period is over. The metrics there are about cost reduction, and they always need to be business metrics rather than technology metrics.

Are you seeing a rise in the use of containers as well?

They’re definitely a standard or common part of the thought process now about how applications or services are delivered. There are lots of solutions and many different vendors but it doesn’t seem like we’ve got a complete handle on how to do it at the moment. They will be part of a big push moving forward, but I’m not sure how many people are going to be rolling their own containers and how many are just going to be using a pre-existing service which on the back end may or may not be containerized.

What about volumes of data? Do you see the companies you work for having a strategy to move from gigabyte to terabyte to possibly petabyte over time?

They’re starting to think about it and the starting point is usually why they’re paying so much for data storage. For many it’s because they’re backing up whole systems and keeping them for seven years rather than just backing up the data that has business value. Fortunately, it’s a little easier than it was ten years ago to get people to think about classifying data. The need to separate payment card information has been a big driver, and people are now wanting to classify Personally Identifiable Information (PII) as well because they’re a lot more aware of how a data breach will be all over social media. They’re also starting to think about the value of data, and value comes in a range of flavors, from compliance to broader analytic trends.

Does DevOps come into the picture as well? What does it mean to you?

DevOps is about making sure you can get a service or product into the wild so that it can be generating value, which then gets fed back into further development. Whatever segment you’re in, the customer market changes much more frequently than previously, so you need to be able to get an idea developed and out there so you can take advantage of it.

How do you know if DevOps is the approach for your team or the project that you’re working on?

It depends what you’re doing. If you’re developing something that customers interact with that needs to respond to a changing environment and doesn’t require a large capital outlay, then DevOps is almost certainly the way to go. I still struggle with the idea of people trying to run solely infrastructure projects in a DevOps way. How can you get into that do, learn, update cycle when you need to know all the requirements upfront because you’re spending millions of dollars on a big bit of tin? It’s often really hard to reconfigure tin, unlike code which can be refactored and modified.

What’s your one top tip for someone starting out as a Solution Architect today?

I guess two things. Firstly, trust yourself because you’ve got this far by showing skill with one or more technologies and you’ve demonstrated you can solve problems that are outside of the technology space. The other thing is listen and keep asking questions like what does it mean to you, what advantage will it give you, and what limitations will it let you overcome.

What’s the next big thing that will change how we work?

I’m actually really interested to see how the office environment changes when we all come out of lockdown, because we’ve proven that we can work very successfully from home. I talk a lot about being able to talk to people and ask questions and get opinions, but how does that work when you’re not in the same room? So I think that’s going to be the next major force which shapes the office environment and the business environment, but how that’s going to play out in a technology sphere or the people arena I don’t know.

Chris Slee is a Brisbane-based independent Solution Architect who specializes in helping clients make a seamless transition from where they are now to where they want to be. He has worked with government agencies, corporates of all sizes from garage-based to multi-nationals, and not-for-profit organizations, moving and improving their information infrastructure, integrated applications and data assets. Chris is keen to share this range of perspectives with other to enable them to make better technology decisions. He’s also the principle behind LongEdge Press which publishes translations of French fencing manuals and LongEdge Fencing where he teaches Renaissance swordplay.

Look out for other episodes in this series, featuring Michaela Murray interviewing:

• Darren Dawson, Pre-Sales Architect at WARDY IT Solutions
• Kris Bock, Machine Learning / Data Science Architect at Microsoft
• Lace Lofranco, Senior Software Development Engineer & Architect at Microsoft

This series of interviews concluded with an online panel discussion hosted by Michaela, during which all of the Architects discussed what the future holds for IT Architects when there’s no blueprint. The recording of the lively and informative debate is now available to watch on-demand.