Y-DATA student during the day, freelancer at night
One sunny morning we sat with Franz Krekeler for a conversation over Zoom. Franz is currently a student in the Y-DATA program in Tel Aviv, but he is so many things, read on to find out!
Let’s get to know you for our readers. How would you describe yourself professionally?
I am a freelance full stack developer. My current tech stack is Python, React, TypeScript and NodeJS. I work with the AWS Cloud, NoSQL databases, SQL databases, and a bit of DevOps.
How did you get into programming initially?
Four years ago my dad had his 64th birthday and I wanted to surprise him with a self-built photo-booth for his party. I found a script online where you connect your camera and printer to a Raspberry Pi. The script was written in Python, and then I was like “Oh, this is so cool!” I tweaked the script, built the photo booth and had so much fun. So, I got into programming out of curiosity.
How did it happen that you could just tweak the script? Did you have a relevant background?
When I was 14, a friend of mine called me and said, “Hey, look, there is the video we’ve made!”, and showed me our video, which was featured on YouTube front page with almost half a million views. I was really impressed. After school I started a career as a freelancer, working as a cameraman and producer. I worked for large companies which needed an internal promotional movie or even a commercial. This line of work allowed me to be very free at a young age and travel around the world. At some point, I worked at Der Spiegel, Germany’s largest news magazine.
Initially I began to experiment out of curiosity, and I built my own banking system for my taxes through Google Sheets, including expense category analysis and more, and then I realized I should get deeper into programming.
I found out about BA degree in human-computer interaction (which is a mixture of 50% computer science and 50% psychology) at the University of Hamburg. For me it was not an immediate career shift. I basically locked myself into my room for one year, and I was only coding. Day by day I found myself immersing into another coding adventure.
Looks like you have a very impressive list of professions and interests. What was the most helpful skill that got you enrolled at Y-DATA?
In summary, it was my interest and my passion for Python and numpy.
At one point I was in Munich during the Oktoberfest and instead of going to the Wiesn I decided to finish reading the book — Python for Data Analysis: Data Wrangling with Pandas by Wes McKinney, the founder of Pandas. During the Y-DATA entrance test I had a lot of fun with competitive programming.
What are your reasons for choosing Y-DATA? There are a lot of ML programs around the world and even a lot of programs in Israel
By coincidence, I hired a freelancer on Upwork for a data problem and he told me about the Yandex data school in Russia (YSDA), where he had studied. I have heard of Yandex company and I read about this program, and I was like, “Should I go to Russia and become this crazy data scientist?”
Then I found out that Y-DATA started in Tel Aviv, I applied and then I got my visa. YSDA has a long history of over ten years in Russia, but there it’s longer — full two years. In Israel it has a format of only two semesters. It’s not a long time, so I was excited to try it out.
What did you feel you were lacking before entering the program?
Before the course I worked mostly by trial and error. In my data science projects I tried out different classifiers. What I was looking for is that person who would take you by your hand and show you: “This is the field you are in. This is what you can do and this is how you do it.” So what was missing for me was a mentor to guide me as I was getting more and more into data science.
Does Y-DATA provide you with this?
Definitely. What really, really got me excited was being surrounded by like-minded people who are curious, who like to solve problems and come from different industries. It feels like a good environment to grow.
Who is your study partner?
My study partner is called Omri. On the first lesson we were sitting next to each other. And it was funny because he did a semester abroad in Katzhütte, Germany at the KAT. We could relate a lot through his experience as an exchange student and my experience now as being a foreigner here in Israel. He’s a local and he has helped me a lot with questions I have in general. He is a very helpful person, and we were just randomly sitting next to each other.
Looks like being a freelancer from a young age made you learn how to organize your time and how to be self-reliant. How would you advise others to learn from your success?
If you want to become a freelancer in the field of data science, it is very important to do it for the right reasons and not because you read somewhere that it is fun. It takes a lot of self discipline. It takes at least a year until you can really make a living from it. My advice when you start, don’t jump immediately into the pool of sharks — take it slowly.
At first, take a part-time job until you go fully into freelancing. Also, you would need to jump a little over your shadow — if you are used to a full-time job — out of your comfort zone. It’s unlikely things will automatically fall into your hands without proactively searching for opportunities.
What are the differences to a regular university?
What was the most pleasantly surprising thing about the program?
I was surprised to see how well teaching was organized — both offline and online: to be able to follow up on a lesson later through the online learning platform and to hand in the homework. We have a full-blown digital classroom as well — not only lectures. Now with the coronavirus, it is really useful and fully functional: one student went home abroad and he still participates in the program online. My university is still lacking behind light-years in digitalization.
Another pleasant surprise were the guest lectures by high-tech CEOs and CTOs on a very high level.
The classmates are extremely liked-minded and come from diverse backgrounds. From startups to well-established large companies.
What are these industry projects?
For Y-DATA I take part in two projects. It is called industry project and you get the chance to work with companies all over high-tech. In my case one is at Anodot company, which is anomaly detection in time series data. And the other one is at NLP startup in the legal domain, which is building a web platform for searching and analyzing law documents. In the first project, we have an interesting team: among others, a SAP consultant and a post PhD. So, we’re a great mix. The second team is smaller: it’s only two people, so you can work closely together.
In your next projects are you going to freelance your skills as a data scientist?
I already worked with cloud infrastructures. One is actually my first freelance job here in Tel Aviv — setting up Elastic search on AWS.
Depending on demand, I would like to solve the problems of logistics and optimization. I already experimented with genetic algorithms. One of them is to simulate dog crossbreeds. Right now I am fully occupied as a full stack developer but I am open to opportunities even if it is going well as it is.
What is the most challenging thing for you in Y-DATA?
The Russian standard of the amount of homework that needs to be solved in a few weeks. Statistics part was challenging too, even though I have a statistical background. Also the high amount of math skills required. I was fortunate to have a partner who helped me to keep up with the high pace.
How is the learning experience similar or different from the University of Hamburg?
That’s a very good question. Sometimes, I have a feeling that the amount of material covered through a whole semester is covered in Y-DATA in one week, it is so intense.
Another huge difference: in Hamburg University we had a very chaotic digital platform. It feels like the university still struggles in the digital world. The platform’s usability was lacking, it was hard to perform even the basic tasks, like to download the materials of the course.
In Y-DATA it is was a total relief to discover the accessibility of teachers and students.
In the university setting you don’t get any practical experience. Now in Y-DATA, I get a lot of practical knowledge and problem solving experience connected to the industry. You have so many people around you who work in the field and solve real problems. Few universities have that.
What kind of problems would like to be able to solve once you are finished with the studies?
I can give you an inspiring example: Lior Wolf, a professor in Tel Aviv University, also from Facebook AI research team. He led one of our research seminars. He has authored quite a list of papers in different fields. This is the kind of prolific creativity that I want to learn from! In one of the papers he generated images to count repetitions in images. That is something — I didn’t have an idea that it was possible. This is a creative solution for a creative problem. I would like to learn how to make simple but effective inventions in the field.
You started your programming career 4 years ago. Where do you see yourself in 4 years time?
I started out of curiosity — with the photo booth for my fathers’ birthday. In four years I would love to teach — I hope I would have the credibility — and pass on the knowledge and help the people who want to become a freelancer and a data scientist.
