Breaking into cloud in 2026: A case study for new engineers

Overview
Getting into IT from scratch is harder than it used to be, and the Australian market is in a rather strange spot right now. The path I took to get into the industry was once a completely plausible one. It’s still technically possible to follow, but it’s far less common today, and AI certainly isn’t making it any easier.
My own start was a little unconventional on paper. I left school at 15, did hospitality for 3 years, took a break, and then started my IT career at 22 on a helpdesk. Along the way I chased vendor certifications, passing some and failing others. They had little bearing on getting my foot in the door, as the technologies I trained on weren’t ones I used immediately. I remained at level 1 for 5 years, taking phone calls and doing basic tasks. However, during that time, I was also focused on career progression. For me, that came in the form of studying, networking (with people, not routers and switches), asking questions and forming relationships. Eventually I moved into infrastructure engineering, which has since evolved into platform engineering or DevOps engineering depending on how you look at it. In a previous blog I talked about how I was “conditioned” like others to grow this way.
That was normal for anyone who started in IT 10 to 15 years ago. The helpdesk was the on-ramp. You learned by doing: touching real systems, fixing real breakages and working your way up. That suited people who learn best through hands-on experience (kinesthetic learners) and didn’t require a degree to get started. A university degree can be a good springboard into graduate positions, but it was never the only pathway. The problem now is that those entry-level rungs are thinning out, and the hands-on learners are the ones who stand to lose the most.
So what does it take in 2026 to move into the platform engineering* space as a newcomer? To find out, I spoke with someone who has recently come through it and asked him a series of questions. He’s asked to remain anonymous, so I’ll refer to him as Joe and keep the details light.
* For the sake of brevity, I’ll use the term platform engineering to describe those who work in the cloud, platform and/or the DevOps space.
Meet Joe
Joe’s way in was different from mine. I started on a helpdesk and worked my way up, whereas Joe went through university and into a graduate role. I asked him about his background and experience. Here is what he shared:
- Completed Year 12
- Completed a bachelor’s degree in Computer Science, which included a cloud computing module with hands-on AWS exposure
- Entered the workforce through a one-year graduate role, with a heavy emphasis on mentorship and shadowing, which evolved into a full-time position
- Currently has 3 YOE in the IT industry
- Currently works as a platform engineer in the finance sector
Joe was upfront that he disliked studying and struggled in his first year of university. He also mentioned, to my surprise, that his impression was the university deliberately enrolled first-year students into an unnecessarily high number of subjects to weed out those who weren’t serious. I can’t say if this is a common practice, and it is only one person’s view, but I was baffled to hear it. University is a significant financial commitment, and that seems a rough way to find out who is committed. Still, Joe pushed through and finished his degree.
Why platform engineering?
The short answer is that Joe more stumbled into platform engineering than had a strong desire to pursue it from the beginning. Platform engineering is a space that usually requires a background in either infrastructure or software engineering. While jumping straight into it from university is possible, it can be a rather daunting and complex activity. Still, as Joe has highlighted, there can be exceptions to that rule.
What drew him towards it was that he liked backend solutioning. At university, a project required backend knowledge and tinkering. In this context, backend was not simply the software backend (like most of us would interpret). It was everything backend, including the infrastructure portion. That project was part of the cloud computing module mentioned above, and it gave him his first real taste of what using cloud services was like.
Software engineering was still his first choice, though. I think this makes a lot of sense, as when most people think about computers and IT, they think software. I suppose the saying “ignorance is bliss” rings true here, given that in order to run that software, a lot of other components need to align. Unfortunately for Joe, the responses from software engineering jobs dried up. He was far from alone, as the Australian market currently has plenty of graduates and juniors competing for a limited number of roles, and there are many factors behind that.
So Joe did a YOLO move. Over about four months, he put through roughly 50 applications to companies in the platform engineering space. As luck would have it, he included a cover letter with just one of them, and that was the one he heard back from. That application landed him an interview, an offer and, finally, a position: a one-year graduate role, which then evolved into a full-time position.
From here, I asked Joe if he was still enjoying the space. He said that as he takes on more responsibility and sees the bigger picture, he is enjoying solving technical problems. To add some commentary of my own: for those of us in IT, we start small by addressing isolated issues here and there. When we get more experience and our horizons broaden, we move on to design work and architect solutions with multiple components. It requires a broader level of skill and expertise. Seeing all of this come together is gratifying, and Joe is clearly getting a lot out of it.
AI and how it is affecting the landscape
AI is a major disruptor, and it is changing both how newcomers get into the industry and what the work looks like once they’re in. I asked Joe about his thoughts and how AI has affected his career, from applying for roles through to the work itself.
What Joe said
Joe told me that when he first started looking around, he got wrapped up in the hysteria around AI and the idea that you needed to know it to be effective. In essence, he was worried like everyone else about whether the junior roles would get replaced. For him, it was more about AI potentially impacting junior roles than it directly affecting his own job search.
We also discussed how, once the technology is understood, it can become a very powerful tool. Joe raised the way juniors and grads use it. The risk, as he sees it, is using AI to just get the answer without understanding the how and why.
My take
While it is true that junior roles are drying up, what is also true is that eventually seasoned veterans will start to age and move out of the workforce. When this happens, what do we do? Do we simply accept that juniors coming up the ranks may have knowledge gaps which we hope AI will fill? The helpdesk was where I learned my fundamentals, and that is exactly the rung that is thinning out.
Having confidence in AI is warranted, but it will not replace engineering outright (I dig into why in The Jevons Paradox). Fundamental skills, conceptual knowledge and technical understanding are still vital. I personally don’t write much code anymore, but I still review plenty, and you can’t review what you don’t understand.
At the time of writing this blog, recent news makes the point. Nvidia has changed CEO, but what the company is doing hasn’t changed: it is building the chips used to train ever-better models. However good those models get, they won’t replace engineers outright. And Microsoft had a global outage that ran for well over 12 hours. I’d wager it wasn’t agents that fixed it, it was people. AI can help us build and run systems, but when those systems break at scale, it is engineers with solid fundamentals who diagnose and restore them. That is exactly the skill set a junior needs to build.
AI has been likened to lightning in a cage: enormous power, but only useful if it is contained and directed. As a user of AI almost daily, I can tell you from experience that it goes off on tangents. It also lacks business/domain understanding, so it relies heavily on the prompter to fill the gaps. It needs direction and correction to get the desired outcome. Otherwise, you’ll end up with a lot of generated content that the AI will make you believe is exactly what you need, when in fact, it is not.
Which brings me back to juniors using AI for the answer. Often we just want the answer. We don’t care how the answer is determined, we just want it. Figuring out a problem or addressing a breakfix issue is ultimately the goal. But what happens if we don’t understand how we got there? What happens when instead of us steering AI, the AI steers us? Sure, perhaps we do get to the resolution. We also have no idea on the how. This is a major problem that will plague the industry for some time until the whole landscape is shaken out. It will cause a cognitive drop in the workforce, and it will leave us not understanding vital information that is key as an engineer.
Advice and recommendations
What Joe said
Towards the end of our conversation, we discussed advice for others. I asked Joe what general advice he would give to grads entering the IT industry. This is what he came back with:
- Take it slow and learn first principles. Don’t take the easy way out by rushing for the solution.
- AI is a helpful tool, but you should drive it, not let it drive you. If it does everything for you, you will learn very little.
- Don’t be afraid of failure.
- Don’t worry about not feeling valuable within the first 1-2 years of your first role. Accept that you won’t produce good results while you’re learning, and that’s okay. Focus on learning and contributing in your own way.
- Have a growth mindset.
In the same vein of advice, I asked Joe if he felt that being a platform engineer at a grad level was an effective choice (more on why I asked this below). Here is what Joe said: “It depends on the metrics for success. If you’re trying to target closure of tickets then a grad can be an effective platform engineer. But if you are targeting solutioning then no way. There is not enough experience at this level.”
He explained that his point on ticket closure was about engineers who just want to close a ticket to look good against things like KPIs or SLOs, without considering what the “proper” solution should be. Joe has since moved on from his grad role into a full-time position, and as his experience grew, so did his mindset for creating solution designs.
He went on to say that he personally benefited significantly from shadowing others (watching them work, asking questions and learning). Good mentorship has allowed him to progress at a good pace (faster than I did).
To wrap it all up, I asked Joe a final question of whether he would recommend to others how he entered the IT workforce. He said, in a nutshell, that it depends on your learning style. For him, the pathway of university, then a grad role with a heavy amount of mentorship and shadowing, allowed him to prosper. He also advised not to be afraid to ask mentors dumb questions and pick their brains to understand their logic, as understanding how solutions break down is key to success.
My take
The piece of Joe’s advice I would specifically call out is learning first principles. Fundamentals are critical in this new day and age. If you don’t understand what AI is producing, in the slightest, are you ready to take accountability? I certainly wouldn’t.
The reason I asked whether a grad can be an effective platform engineer is how much the landscape has widened. Infrastructure roles 10 years ago mainly consisted of systems engineers focused on Windows on Intel (Wintel), network engineers focused on networking equipment and configuration, Linux/storage engineers focused on Linux ecosystems and SANs, and cyber security engineers who would focus on security tooling. There would be others, but this is generalised. Nowadays, a platform engineer is expected to know all of this at an abstracted level (no need to configure SANs or the equipment in a DC) while also understanding DevOps fundamentals and concepts.
Joe’s answer on learning style echoes what I said at the start about hands-on learners. The helpdesk was my on-ramp, and a mentored grad program was his. They are different routes, but the common ingredient is learning by doing, alongside people who know more than you.
On the topic of mentoring, I have rarely met someone who is knowledgeable and who doesn’t like to mentor. Most people in the IT field (based on my experience at least) are more than willing to show you how things work and all that they know. All that they ask for is your attention while they go out of their way to bring you up to their understanding.
Conclusion
What stands out most to me from my conversation with Joe is how little of his path was planned. He wanted software engineering, stumbled into platform engineering when those doors closed, and got his foot in the door through one cover letter out of around 50 applications over four months. After 3 years in the field, he’s doing really well. It is a good reminder that effort counts for a lot, but it isn’t the whole picture. The market, timing and a bit of luck all play a part.
Joe and I also took very different pathways into the same space. I started on a helpdesk and worked my way up over many years, whereas Joe went through university and into a mentored grad program, and has progressed faster than I did. Neither is right or wrong. So what does it take in 2026? From where I sit, it’s access to a mentored on-ramp to replace the helpdesk rung that is thinning out, a solid grasp of the fundamentals rather than leaning on AI for the answers, and the persistence to keep applying in a crowded market.
I’ll admit I have a bias, based on what I’ve told others in the past. IT is a field where you can push as hard as you want to achieve success. If you want to progress faster, push yourself to learn and challenge your abilities. The harder you push, the quicker you will progress. That is what I have both observed and lived in my own career. But it is entirely optional. You can just as well take it slower, as per Joe’s advice, and for some that is what they are more comfortable with.
Finally, thank you to Joe for sharing his story so openly. I hope it helps someone starting out on their own pathway.
Cover image by aistudio.google.com