From coding the first Kiwi.com algorithms to “But does it scale”?

Posted on 07/09/2020, Teodora Stojšin

Back in 2012 Jozef Képesi co-founded Kiwi.com (formerly Skypicker) and created the first functional version of its search engine.

Many things have changed since then, the company grew from two to hundreds of people and has been to some ups and downs. JK, as everyone in the company calls him, is 8 years wiser but when speaking to him, you still feel the startup vibe. Learn more about him and his approach to leading the engineering team of one of the fastest-growing Czech startups.

 

What was your first job?

Before I joined Kiwi.com, I worked as a freelance developer, mostly in what you would call today Big Data and NLP. However, back then we didn’t use fancy names — I was playing around with a bunch of data.

 

Did you always want to work in IT?

No, I actually wanted to become a structural engineer, I was always super fascinated by physics and one of the main characters in Prison Break, which was a thing at that time, Michael, was a structural engineer, which inspired me. I nearly enrolled at Bratislava University to study engineering, but the university in Brno actually accepted me to study IT before I even had a chance to do the tests in Bratislava, so I just went with the flow and ended up in Brno, where I met a lot of amazing people, including our future co-founder.

I studied Informatics and Economics, mixed with some law. Basically, I studied anything that could get my attention. Unfortunately, if I am not totally immersed in something, my attention span shortens and as soon as I don’t see any progress, I’m looking for the next thing to get my interest. I tended to drop subjects after a year or so, but looking back, I was always able to get the best out of every field and I don’t regret not finishing any of my studies.

 

Explain your career path. Did you take any detours?

I did a lot of random stuff — such as working for a call center or waiting tables — during my high school studies, but co-founding and then leading Kiwi.com was my first full-time (or better 24/7) job.

 

“Strive for impact, not perfection.”

 

What type of CTO are you?

A very pragmatic one. I believe I can find simple, effective solutions to problems that seem too complex at first sight. I sharpened this skill pretty well during the first years of the company when it was our only way to survive.

I can be very impatient and sometimes ignore the details but it is helpful when someone comes and spends ages trying to find the best solution possible. In these moments the important thing is to sit down and say “What we have is good enough, let’s move on!”.

I believe my mindset helps with that.

 

Which emerging technology are you most excited about?

There is a bunch of stuff going on nowadays, but tech comes and tech goes, what was cool today can be dead tomorrow — just look at the evolution of JS frameworks during the past few years. Huge players like Facebook had to come in and tell everybody “This is the way how you build stuff from today, just deal with it and go think again about your brilliant *let’s change everything again from scratch* ideas”.

As an engineer, you need to realize no customer is ever going to see a single line of your code, so I ask everybody to strive for impact, not perfection. The only people reading your code will be your colleagues and they have to react with “Ok, I can get this.” instead of “Wow, this looks awesome, but I have no idea what it does.”

At the same time, we all need to realize there will be a huge change, which is emerging now and the break will happen once everybody will be able to create their own tools and build businesses without knowing any programming language.

I mean — there has to be a point when software engineers become nearly obsolete and computers will just build most of the stuff based on our requirements (or even generate the requirements based on the needs of our civilization).

Basically, the profession itself evolves to a peak where it will become a lot less relevant, or dies.

 

When do you think this happens?

Fortunately, as I still love coding and as we see that today’s AI might be hitting a blind road, I don’t see this happening on a global scale in the next decade, we are really just doing the first baby steps and the first tools like this are tasting the market.

 

Are there any technologies that you think are overhyped? Why?

If I should pick one of the many buzzwords, it would be cloud and the “digital transformation”, which are often seen as a way to fix literally everything inside your organization.

In reality, moving to cloud-based systems doesn’t mean your scaling problems are going to magically disappear, it just removes some of the already known obstacles and at the same time introduces a lot of new ones, which you have never seen before, but now the big guys are backing you.

 

What is one unique initiative that you’ve employed over the last 12 months that you’re really proud of?

I can’t pick a single thing we did, as this is not the way we operate. We have multiple ongoing initiatives, each of them being a never-ending, evolving process.

The enterprise is like a living being — the world and its conditions are constantly changing, so it has to adapt & evolve to survive and improve with every single iteration to make sure there is no competition taking over.

This is a double-edged sword, though, as you can never feel that the job is done.

However, on the other side, an organization that is able to realize this and keep this attitude in the very core of its DNA is going to be evolving every day, getting stronger and more impactful. This is how we operate at Kiwi.com, ever-evolving, testing, innovating, learning, and moving on at a fast, rewarding pace.

 

“This is how we operate at Kiwi.com, ever-evolving, testing, innovating, learning, and moving on at a fast, rewarding pace.”

 

Even though you mentioned digital transformation is over-hyped, are you leading something that resembles that process? If so, does it emphasize customer experience and revenue growth or operational efficiency? If both, how do you balance the two?

This is actually very hard and I don’t think you can simply pick one or the other. But very pragmatically speaking, for us as engineers, cloud or digital transformation means better operational efficiency, as it enables focus on the actual product, which leads to better customer experience, which ultimately leads to growth.

 

How do you align your technology use to meet business goals?

We always leveraged technology to support our business needs, not vice versa. First, there has to be a customer need backed by a business case, only after that, we look at how we can build/support it.

We often saw engineers going off-road, chasing new technology, forgetting to ask “How does it impact the customer?”.

Once you have a good answer to that question, I am pretty sure we are able to build a business case for the majority of the cases.

 

Do you have any trouble matching product/service strategy with tech strategy?

Both founders being tech guys, we were born as a tech company, so this was never an issue.

 

What makes an effective tech strategy?

The KISS (Keep It Simple, Stupid!) attitude is very efficient.

 

What predictions do you have for the role of the CTO in the future?

Defining the role is quite tricky, as the CTO in a startup is usually also one of the founders and the role evolves over time. It starts with being the Chief architect, moving through Chief recruiter and Chief coffee brewer for the recruited team (as if you did your recruiting right, they know better how to code than you) to the guy who knows stuff”, leading him to constantly ask “But does it scale?”.

Through the whole process, the main function of the role should be about the principles of the company culture.

 

What has been your greatest career achievement?

In the early stages of the company, our first engineering team literally disappeared in front of our eyes because I had no idea how to manage them. We found one of the members actually left for a job in McDonald’s, so it couldn’t feel worse. Looking back at this and realizing that everything is about the team, the greatest achievement is knowing that we have such a great team that I believe we can actually achieve anything.

 

Looking back at 20:20 hindsight, what would you have done differently?

Nothing and everything. We wouldn’t be the same without all the mistakes we made. On the other hand, I am afraid what could we become if we didn’t make them…

 

What are you reading now?

Andy Grove — High output management. It’s fascinating to see how nobody figured out anything better than the principles used in Intel years before today tech leaders like Google even existed…

 

Most people don’t know that I…

Read books & speak Hungarian…

 

In my spare time, I like to…

… travel, travel, travel, party and meditate.

 

Ask me to do anything but…

Some people will disagree here, but never ask me to run a city marathon. I nearly missed my flight once because of such an event and since then I never understood the concept of locking down historic city centers and blocking daily people’s lives, so that some people can run on the streets, while there is a stadium a few miles away, created for that purpose. It makes no sense to me.

 

Editor’s note: The interview was done back in 2019 and based on interview questions by IDG CTO Sessions feature.

Read next