For me, it’s easy because my career path was clear already in elementary school
I went to school in Sweden, where the financial situation of schools was on a completely different level compared to Finland at that time (early 90s). Concretely, this was evident in every classroom having a desktop computer available for free use, and I certainly made good use of it. MS-DOS with its commands became very familiar, so I could run Gorillas, Ski or Diet, Skyroads, and others
Then, in 1994 or 1995, the school got a T1 internet connection (1.5 Mbit!), which at that time was like having something like a 12.5 gigabit internet connection today. And then in 1997, we got about 15 (thick) laptops per class (class size was around 10 at that point…), also for free use. We also had coding lessons for everyone, with QBasic as the language.
I started high school, but then I heard from a friend that IT could also be studied properly. Discussions with the guidance counselor and then straight to vocational school for a data processing qualification (datanomi) in 2002, to learn more about installing servers, assembling computers from components, and surprisingly a lot of coding (C, Visual Studio .NET, Delphi, ASP.NET), and also several accounting courses and business courses as part of the education, touching upon investing. Although data processing specialists (datanomi) are often criticized, I found that to be a very high-quality and comprehensive education - perhaps there are regional differences? But it was clear, however, that a mere data processing qualification (datanomi) wouldn’t be enough for me 
In between, I already used my skills for small-scale web development for money, for local organizations and small businesses. That was good practical training too, even though the compensation for it was mainly in the range of expense reimbursement
The more important part was honing my skills in real projects.
After the army, it was time to go to a university of applied sciences in 2006, to get an IT engineer’s degree. There, the choice was practically “hardware for Nokia” or “mobile software for Nokia”
Well, not really, but Nokia’s “stamp” was very clearly visible in the themes of the specialization courses. I chose the software side as my specialization, where we then coded with C, MatLab, Java, PHP, C#, and did SQL, physical electrical circuit connections, and embedded systems for AVR microcontrollers, especially ATmega128, less surprisingly. And we read microcontroller documentation day after day - good training for reading quarterly reports on the investment side 
And in my free time, of course, a hell of a lot of coding, especially for the web
At this point, I already had my degree in my pocket in 2010, and it was time to start earning money with it.
My first job was then found far away in Southern Finland, where I did consulting work for 5 years starting in 2011, and in line with the school theme, the language selection was already comprehensive at this point, as my first position involved using Java EE, C#, HTML, CSS, SQL, and Rational’s toolset. And in addition, in my second position, I got to do work for the state (too), at a Unix level! Shell scripts and commands became very familiar there
And the second part involved B2B messages using EDIFACT, SAP IDOCs, and XML. Alongside that, I also became familiar with companies’ ordering and invoicing practices, as we investigated errors in data, even in connection with Finland’s largest companies.
Well, everything in its time, and the desire to do something new took over, so my first job change was from a large consulting firm to what was then still a startup, now a publicly listed company (which seems quite popular here). There, the first task was to assemble Ikea office desks as a team! Then laptops on the table and to work
Here, I got to implement Python, which I had studied in my own time, at a practical level. A successful investment earlier! Python became one of my favorite languages, even after long-term professional use. We then did projects extending to Germany, where we made several longer trips. At the same time, we also started developing our own auxiliary tools, and then I researched what would be an excellent choice for a frontend tool. Ultimately, React became the choice, which at that time (~2017) was not yet the giant it is today. But the steps were clear even then, and I’m not surprised by its current popularity 
Then megalomania struck, and in late 2017, I decided to test my mettle in one of Finland’s probably most technically demanding companies. And it was exactly that! I got to learn an enormous amount of new things every single day, which was a real goldmine for the future. The main tools were AWS, Node, and the familiar Python. From there, my own enthusiasm for Node and later TypeScript took off, and they are still my favorite languages, where everything is in a good balance - not too bureaucratic, but not too wild west, and at the same time, there’s enough power. And work, in the big picture. However, eventually, I started to be drawn more towards full-stack, instead of pure backend.
So, nothing else but a new startup in sight in 2019! This time, I got to dabble with blood
. Literally. The company’s business is blood analysis, and my part was to develop the ordering system and the analysis tool itself, which the company’s doctors (literally!) then used in their analysis work. In line with the theme, new languages (well, fortunately not literally anymore
) were added to the toolset, with Go and Angular as the stack, in addition to the already familiar PHP and React. However, the most important lesson came from AWS and Docker, both of which were then key in my next job.
Actually, the only reason for changing jobs in the autumn of the same year, 2019, was that I got my dream job, and the position was in the sexiest thing of all - user interfaces in airplanes! Besides the ones you might already be familiar with, there are also loooots of others. Here, I got to travel around the world
for work, and California became familiar
Since we were tinkering with “all” systems, the toolset was accordingly: Qt QML, JavaScript, Node, AWS, Docker, Jenkins, Linux, TypeScript, Serverless, SQL, SQLite, React, WebSocket, XML, JSON, and I probably even forget something
Easy as pie. But we got it finished and a Red Dot in our pocket. Then, making more tools, and the stack didn’t shrink much at all, and work continues even after all the world’s crises…
So, what can be learned from this novella? Well, at least I’d say it’s not worth getting too stuck in one or two technologies if you want to experience all sorts of things in your career
And as for career development, it’s a stark truth that salary develops best by changing jobs at appropriate intervals. Perhaps every 2-3 years, until the salary starts to be sufficient. What that is, of course, depends on you and the industry.