I was born in 1988 in Neustadt an der Weinstraße, Germany. After finishing School in 2007, I studied Computer Sciences in Darmstadt and graduated with a Master's degree in 2015. After working for 3 years as a consultant and software developer for nterra and eXXcellent, I decided to quit my job and start working as a freelance software developer. Together with two of my best friends and colleges, I founded

CodingGene

Today I am still making a living as a freelancer, but in Spring 2021 we founded Tyromind, and took our first steps into the exciting world of product development.

How I learned what I know today

I started programming when I got my first computer at the age of 14. I was immediately drawn to the web, so I initially taught myself HTML and CSS. Although web standards were a real mess at that time (IE 5.5 🤬), I had a whole lot of fun learning and developing my skills. It didn't take long until I noticed, that static web pages are pretty useless in many cases, so I learned PHP and MySQL to make my pages interactive. At that time broadband internet was a real rarity in Germany, so I had to use a 56k modem for connecting to the world. Often it was frustrating, but it very early taught me the importance of optimizing pages for loading speed. At that time I also had my first contact with JavaScript, but I really disliked the language, and it would take the rise of NodeJS until I gave it another look. In fact, for many years, I was actively advocating against the use of JavaScript in web pages, except for purely cosmetic purposes.

I stuck to this technology stack until I got into University, where I was first introduced to statically typed languages like C++ and Java. It was love on first sight. Not that I wasn't interested in other languages before, but I always evaluated technologies by whether I was able to use them for one of my web projects. At the time I couldn't afford to spend money on servers for running a JVM, and web-space with PHP/MySQL was essentially free. However, it took a few days of learning Java until I realized, how much a statically typed language can help maintaining a large codebase. I really wanted to not just learn it, but actually use it, so after one week of going to university, I applied for a job as a student software developer at Fraunhofer SIT. They hired me, and even more importantly gave me access to the resources, as well as the freedom I needed, to dive as deep into the Java universe as I wanted.

I stayed at Fraunhofer SIT until I got my Bachelor's degree in Computer Sciences in 2012. It was a great time, and sometimes I really miss the way I used to work there. Not only did I have all the freedom I needed, but I also learned a lot about applied IT-Security, how to write scientific texts, and met a great number of highly talented people.

However, after 3 years I felt that it was time to do something new and exciting. I wanted to get a taste of startup-air, so I applied for a job at CoSee. And it was worth it. I never again learned so many new things, so many new concepts, tools, and technologies in so little time. From the first day, I was working with a team of young, highly talented software developers. I thought I already knew many of the tools they were using (such as Git for example) , but I quickly realized how much more there was to learn, and so I did. Over the next year, I worked on my first Android app, dived through a huge JEE backend codebase, queried Elasticsearch and a massive Oracle database, and last but not least applied my knowledge about web technologies, this time with JSP as a framework.

Unfortunately, after one year of working and learning at CoSee, I was forced to search for a new job. My main occupation was still studying computer sciences at the university, but my parents could no longer afford to pay for my rent, so I had to earn more money.

After looking around for a while, I was offered a job as a student assistant at nterra, a consulting company near Darmstadt. They just started a business unit for Java Enterprise consulting, and so I got even deeper into the Java web and backend world. Meanwhile the web started to discover Single Page Applications and new JavaScript frameworks were being used for frontend development.

During a project led by my university in collaboration with some industry partners, I had first contact with a SPA framework: Knockout.js. To this point I had only ever done server side rendering, but the possibilities of creating a much better UX with SPAs really resonated with me. I immediately started to learn all I could about the world of frontend JavaScript frameworks and build tools. However, being used to work with large codebases, I was wondering how to write big applications with them. So, it took not long until I started to use TypeScript for all of my frontend work, and I never looked back since.

After writing my Master-Thesis, I started working full-time as a consultant for nterra. I was sent to clients to support them with Java Enterprise and Web development projects. Unfortunately, the world of enterprise software development in Germany is quiet slow when it comes to adopting new technologies. Hence I was forced to mostly work with JSF, which in my opinion is the worst web-framework that ever existed. During this time I didn't really get to know exciting new technologies, except for Spring Boot, but I learned a lot about how to deal with customers and clients.

As two years of writing enterprise software had passed, I felt like it was once again time for something new. I craved for new knowledge, and even more for people I could learn from. So I started to search for companies where I could find what I was looking for, and finally I found eXXcellent. Once again I was surrounded by highly talented people. Also it was the first time I got hired as a team-/technical-lead developer. I was assigned an in-house project, and I was more or less free to choose the technologies for implementing it. After some research we started with React and TypeScript for the frontend, Java for the backend and PostgreSQL for persistence. It was a pleasant mixture of professional software development and learning new stuff.

Then came one of these days that turned my professional life upside-down. I was talking to a good friend of mine I knew from my time at CoSee, and he asked me, whether I would like to found a company and get into freelancing. At first I denied, because I had just started working at eXXcellent, and I really liked it so far. However, after thinking about it for a while I agreed. The main reason was, that I always dreamt about having my own company, being my own boss and making my own decisions. Although I always imagined having a company which creates and sells a great product, getting into freelancing seemed like a good first step into the right direction. And of course you can just earn so much more money as a freelancer. To this day I am still unsure if german software developers are generally underpaid, or if freelancers are way overpaid, because of the large gap between the two. In my case my income increased by 450% over night, for doing nearly exactly the same work. So, we founded CodingGene.

My first client, a company that I knew from my time at nterra, let me participate in the modernisation of their customer portal. Everything should change at once, backend, frontend, database. Sounds challenging? Yep, it was. However, from a technological point of view it was just great. It was the first time I got to work with Docker in a client project. We used Ansible for automating deployments, as well as setting up our own infrastructure for CI/CD with GitLab. I introduced the development team to React and TypeScript, which was quiet challenging, since the in-house engineers used to work with JSF for many years, if not decades.

6 months later the project failed completely. I think mainly because creating everything from scratch just took too much time, but I never got to know the exact reason for the cancellation. In any case it opened my eyes regarding something, that every developer thinks or says at least once in their career: "We should rewrite everything from scratch, then it will be much better". Today my mantra is "You should never rewrite everything from scratch, unless there really is no other way." When it comes to modernizing large software systems, you should find a way to do it in small increments, while keeping the system as a whole operational.

That is what we did the next 6 months. We took the old customer portal and injected small pieces of new technology, focusing on improving the user experience as well as the developer experience, without breaking any existing functionality. It worked so well, that we were able to complete several small projects in time with ease, which happened at this company very rarely. After one exciting year I left the project with the great feeling, that I had made an impact on where they were headed in the future.

I already had my next project and client waiting, with lots of work to be done. It was a company based in Australia, which had just opened an office in Germany. At the time they had outsourced most of their software development, but that no longer worked for them. They wanted to transition to in-house development teams, but due to the lack of enough qualified candidates for employment, they needed some external help to get started.

It felt like being part of a start-up. Everyone was fully committed to their product. The in-house developers really wanted to take responsibility, and not only be a replacement for the old external software development company. It was more than a pleasure to help them achieve their goal. Technology wise I got to work with AWS for the first time, wrote some small tools with Go, and worked with an architecture, the old software development company called "Microservices" (although I would call it "A lot of separately deployed services sharing a huge database", which is not exactly what the term Microservices stands for in my opinion). Besides that it was a technology stack, I already felt comfortable with: React , Next.js , TypeScript, and Java, with MySQL, Oracle, and Elasticsearch for persistence.

After working for this company for about half a year, I was assigned to a very urgent project. One of their existing systems for B2B sales needed a modification and extension at its very core. Due to performance reasons (I was told) most of the critical business logic was implemented in SQL, with MySQL being the RDBMS. It was very challenging. We got it done in time and there were no major catastrophes despite the fact, that we made an open-heart surgery on a system we had never worked with up to this point. However, I quickly noticed what a poor choice MySQL was for this system. Whenever I searched for a good solution to a weird SQL problem, I found a great solution for PostgreSQL and a poor/hacky solution for MySQL. I accepted it, because there was nothing I could do about it, but since that time, I am convinced, that PostgreSQL almost always is the better choice when it comes to selecting an RDBMS.

I stayed with this company for more than one and a half year. During that time I also did some small projects for other clients, but they were mostly about consulting and giving advice rather than writing software.

I was now working as a freelancer for 3 years. I had earned a lot more money than before, but I never changed my lifestyle. Instead, I put all that extra money aside, so I could use it to finally create my second company and get into product development. Together with two friends, I started this endeavour in spring 2021. We experimented a lot and again, I got to know so many new technologies in very little time.

We started with a Go and PostgreSQL based serverless backend running on Google Cloud Platform, which we operated through Terraform. Additionally we used React , Next.js , and TypeScript for our frontend, which to this day is my most favorite technology stack. However, we quickly realized, that the backend technology stack caused unnecessary development overhead, which is especially bad if you are prototyping. Thus we searched for a better alternative and found Supabase. It integrated seamlessly with our frontend and gave us an enormous productivity boost. It then didn't take long until we launched our first small SaaS product Scribe Club.

From here on I can't tell you what happened next, because that is the future. I often hear people saying, that you should specialize on a specific set of technologies, that fullstack developers know everything but are good at nothing. I couldn't disagree more. Being one, I am of course highly biased, but so far it served me well. I am convinced that in order to create a great Product, it is not only necessary, but required to have at least one really good fullstack developer on your team. Someone who sees the big picture, but also understands the problems of those, that have specialized on specific technologies.