Per ardua ad astra — and how I ended up in tech.
A short story of how a failed astronomer ended up as a Platform Engineer at ASOS…
Who are you?
My go-to tagline is ‘an avid enthusiast of dabbling in tech’ — but doesn’t tell you much. I’m Dom, a Platform Engineer currently working in Enterprise Data Lake and External Product Feed teams within Data Services at ASOS. I also worked in a dedicated Platform Engineering team (the legendary Team Bacchus anyone?) on various improvements in the data domain.
I’m formerly an astrophysicist and still a deeply amateur sysadmin. I also post on LinkedIn about twice a year — all while saving up to visit Japan with my partner.
Right… so how did you start?
After completing higher education, I graduated with a MPhys (Master of Physics) degree in Astrophysics back in 2019.
I had this outstanding idea to get a PhD, look smart and continue studying asteroids in an area known as Planetary Defence. Having seen whole 20 minutes of Armageddon and having filled out more than a few applications, I received an offer from Queen’s University Belfast (QUB).
While I had some great days performing remote observations and being a de facto data scientist; I quickly realised that a concept of meta-science, the science of how we get results, is more fascinating to me than the history of the Solar System or the distribution of S-taxon asteroids in the Near-Earth Orbit. To understate it, I never quite enjoyed the unadvertised frontline — ‘publish or perish’, among other questionable aspects of academia.
Not being able to save the world from asteroids on this occasion — I indeed perished back in early 2021. After ‘rebranding’ myself as an ‘Independent Researcher’, I looked towards tech instead to find a career. Until that point, I only had a small internship on machine learning and a passion for computers and technology.
This certainly proved to be a plunge.
A bold move — what now?
Around that time, you might have heard that ASOS was setting up a new office in Belfast — the land of Guinness and extra Bank Holidays. With a radio advertisement, I thought there was nothing to lose and applied. A solid week preparing for the interview left me with just over fifty bookmarks… unsurprisingly, I knew quite little coming into this.
But to my surprise it worked. I thus ended up part of the first ASOS cohort of a DfE sponsored programme known as the Assured Skills Academy. At the time it was important development - and personally, there was lot of on the line given my partner (still chugging along in academia) was supporting me. Time was right for this opportunity.
The academy itself was a chaotically packed programme, yet for first time in couple of years, I was excited for learning again. I set my sights on getting a job that linked to my passion for tech and my interest in methodology — DevOps, which after its own rebranding is role itself is now called ‘Platform Engineering’.
Fast-forward through 10 weeks of the academy in a windowless building, and studying for Azure foundation certificates, I was offered a job as an Associate Platform Engineer alongside a couple of my colleagues (one of whom did a blog post herself, check it out!).
How was your ASOS journey so far?
After changing careers, there was a humbling realisation that, again, there was a lot to learn. My previous amateur adventures with old Linux laptops or simple Docker containers in home settings would not get me anywhere near enterprise, or what we call ‘production’ resources.
With a lot of enthusiasm — for the first few months, I would spend few hours after work learning and deploying resources to Azure — first via the Portal, then with PowerShell and ARM templates. I kept adding more topics like basic Continuous Integration/Continuous Deployment (CI/CD) and observability with Grafana. As I did not have traditional Computer Science education like many of my fellow Platform Engineers, the sense of catching-up was certainly there…
Thankfully, my first projects that aimed at codifying CI pipelines helped the domain and let me gain my first experience in bringing tangible value. I soon found my flow while working on background improvements, making often incremental changes such as saving 15% of infrastructure costs in one solution, improving deployment speeds by 20% in another pipeline, or procuring new environments for developers. A lot of these improvements also helped to build relations with peers and other teams.
A supportive environment grew a lot of my confidence back, doing what could be considered of meta-engineering, akin to my adjacent meta-science interests. This in turn resurfaced a lot of my passion which I channeled back doing side-projects, such as a brief appearance back at QUB for my first presentation about DevOps.
After many features completed alongside my teams, evenings working on certifications and many (so many…) development incidents, I worked up to a mid-level over a year later. I then transferred teams into my current role, shortly before being shortlisted for a regional award.
I would be lying if I said that there was no element of self-validation in this — but I hope it is evident that it is possible for anyone to transition. The decision to look at my skill set and pick a career that matched it was a simple engineering problem. It was taking that chance that was the difficult part.
So… what do you actually get paid for?
In practice my day-to-day revolves around developing the technology platform and improving methodology for both the Software Development Life Cycle (SDLC) and developers — by building self-service mechanisms to do so. Between countless scrum meetings discussing ‘story points’, my core responsibilities as a Platform Engineer (PE) can be summarised to:
- Producing infrastructure code, or producing a self-service mechanism to create infrastructure, with Domain Specific Languages (DSLs) such as Azure’s ARM Templates, Bicep and Hashicorp’s Terraform. I would say that half of stories I pick up relate to building new and maintaining existing infrastructure for developers.
- Building and maintaining CI/CD pipelines — alongside infrastructure, CI/CD is core of our SDLC automation and DevOps. Pipelines (a.k.a., workflows) are ‘chains’ of tasks designed to build, and deploy software hosted on infrastructure (like clusters, SQL servers etc.) but also infrastructure itself.
- At my role, I also tend to have a lot of autonomy to support my teams the best I can, while acting as a point of help for infrastructure triage. There is a degree of responsibility in being a specialist (for whatever that’s worth!) knowing that my actions do have a formidable ‘blast radius’. Despite that, this is one of my favorite parts of this job — it's a great catalyst for continuous growth.
- I also prototype (a.k.a., ‘spike’) solutions — sometimes left to the likes of architects. There is a lot of value in exploring new ways to do things in this job — for example, finding out new ways to monitor virtual machine clusters or improving our data storage security. As a PE, I would often get to implement these firsthand and lead the promotion all the way to production.
Anything else you do?
Despite my formal departure in academia, I also volunteer as an IT consultant for an organisation known as FORRT, aiming to develop Open Science resources. We have just received a highly competitive grant to help building up IT infrastructure and instill DevOps culture into the organisation. This is a bit beyond the scope of this blog however…
In my spare time, I enjoy city breaks alongside my partner, ‘homelabbing’, and assembling mechanical keyboards after waiting literal years for parts to arrive (if you know, you know). I occasionally play games on my overpriced hardware alongside friends, for whom I inevitably also administrate servers for. I insist this is a badge of honour, however.
Do you have any words of wisdom?
Maybe, speaking from only a few of years of experience. Take this with a pinch of salt…
- If you are considering going into engineering, my best advice so far is to be curious. This a vast, and rapidly evolving field, so be prepared to put time into self-improvement, especially if you are coming from another field.
- If you end-up in DevOps — you do not need to overhaul systems. Trivial things do matter and stack up sooner than you think. One alert or a dashboard can make a difference.
- As for my ex-scientific/academic colleagues, remember is that you have marketable skills beyond reading or producing papers— occasionally more so than many traditional CompSci graduates — with likes of project management, Root Cause Analysis (RCA), and great problem-solving skills. There is no doubt you would also have some degree of science communication skills under your belt too, which are vital for many tech roles. Your applications (and interviews) should put these front and center, as I did to justify becoming a Platform Engineer. Remember that being a scientist prior switching is an asset — not a detriment.
- Take opportunities. My so-called tech journey started with a spontaneous internship back after my third year of undergraduate studies - fresh after failing several interviews just a year prior. Not taking that chance, or not applying to Assured Skills Academy, would have prohibited me from all those experiences.
‘Per ardua ad astra’ translates to ‘Through adversity to stars’. But even if Platform Engineering involves no stars, by all definitions, I learned that there is nothing wrong in taking a chance — it might just change your life.
Where to find you?
A few links can be found here: kiersz.dev. On rare occasions, I also post on LinkedIn.
