Dan Rachou: Quality is a matter of professional pride

In a world where “enshitification” is spreading, Dan has fortunately found an environment where people take honour and pride in delivering quality.

“Enshitification” is a term coined by Cory Doctorow to describe how many digital products and services slowly but surely get worse as priorities shift from delivering good products to users to securing revenue for owners.

In a world like that, it’s even more exciting to work in an environment where everyone is focused on pushing boundaries and improving what we deliver.

Dan works in Tieto Banktech, which delivers systems to many of Norway’s banks—an area where they are industry-leading.

Tieto has worked with core banking systems for years and has built up expertise few in this country can match—if anyone. I think a big reason we’re leading is that we constantly push the limits of how good the systems can be; we never stop. For a tech nerd like me, it’s fun to work in that kind of environment.

Dan has worked a lot on the NX Core project, or next-generation core banking. It’s a major modernization project about moving systems from the rather old COBOL technology over to Java and Linux.

Transactions are the most fundamental thing for a bank, and the systems that handle them have to perform at a high level. Those of us with a background in low-level programming—where you often work with hardware with very limited resources—have a “performance first” perspective. It’s fun to work on a project where everything else is secondary.

Down in the weeds

Dan has had a technical interest from an early age and deliberately chose studies with a high degree of theoretical teaching. He has specialized in networking, operating systems, and parallel programming, simply because they require a deeper understanding of how computers work.

Proof that I’m a slightly different kind of coder might be that I’m a strong supporter of the keyboard and avoid using the mouse. I also use Emacs rather than VSCode and IntelliJ, and I use Perl instead of Python, which again shows that I may be a bit nostalgic—even if some would go so far as to call me a purist!

Dan lives well with labels like that, and at Tieto he has found colleagues who not only accept his little quirks but also appreciate them.

The safety to fail

When Dan is asked to describe the culture at Tieto Banktech, he highlights the professional respect people have for each other.

I feel very safe with my colleagues and know there’s room to fail without being hung out to dry. We see it as a learning process.

Being open to learning isn’t something only young people have—seasoned veterans who have been in the business a long time have it too.

I have a colleague who has been in the industry longer than I’ve been alive. Still, he’s very receptive when I share viewpoints about code he has written. He considers it seriously, and then he can say whether he agrees or disagrees, or that it’s something we can gladly discuss further.

This openness leads to greater respect, Dan believes.

When such an experienced colleague shows this humility toward me as someone relatively new, I gain enormous respect for him. It also makes me all the more receptive when he provides input for me. There’s no one here trying to put themselves above others—we all share a common goal of building incredibly good products of high quality.

Skills development

To be able to deliver ever better products, it’s important to continuously raise the level of competence. Tieto does this, among other things, through four annual gatherings called CREAM-day (Competence Rules Everything Around Me).

Here, all employees come together to talk shop, and we’re encouraged to give talks to each other.

At the previous gathering, Dan gave a talk he himself describes as quite atypical.

I talked about how I used Perl and Gnuplot for data processing and analysis of performance data from production—an unconventional choice of technology in today’s tech environment.

Both Perl and Gnuplot had their heyday in the mid-1990s. Back then, Perl was the de facto standard for creating dynamic websites, and Gnuplot was the first choice among researchers when it came to making graphs for academic papers. Today most people use Python and Matplotlib, or Microsoft Excel.

The purpose of the talk was to show that “old” technology can sometimes be a more effective and engaging tool to work with. It’s wonderful that we have arenas where we can geek out about the craft together.