EN
There is no easy one-size-fits-all solution to our problem. It's not about being given a clear brief and having a process to get to the result. Nor can you simply pour data into machine learning to solve the problem.
For example, there are giant chains in Arab countries that sell everything from food to clothing. So purely from the data on the name of the merchant, we can't tell what the purchase was. We have to figure out how to better identify the payment by combining a lot of secondary signals and parameters.
It's a lot of creative hacking. We have to figure out how to break the giant problem into subtasks and solve them one by one. That's why we're looking for people who think outside the box and can solve problems that no one has solved before.
You will experience data work from A to Z. You get to be in charge of your project and do everything from initial idea to implementation. Only that way can you learn to think about problems in context.
In other places, you would get input data and use linear regression to get the final number. We prefer more breadth and the opportunity to touch different technologies. Writing an initial script and automating it through Jenkins, for example.
One example is a custom tool to scrape merchant logs from their sites. A colleague made a page in Python and Django that looks up images that look like logos (by size, description, where they link to) on a merchant's URL and offers them for manual review. One quickly picks the right one - and this handy tool has sped up the process about fivefold.
We do work that makes a real difference. Tens of millions of people use our data in their mobile banking app.
It's great that it's a service that's being built in the Czech Republic and used around the world. Only a few companies in the world specialize in this area and we are able to beat them.
In a corporation, the final effect of the work would not be visible like this. Here, a person can invent a new automation script, test it and in two weeks see how it helped on specific data. Suddenly we recognize ten thousand new merchants and the automation ratio goes up by a few percent. And it saves everyone a lot of work. Each time at the end of the year, we can see what a lot of measurable work we've left behind.
We have a team full of talented young people. There are PhD students from the nuclear university or graduates from the math department. Lots of smart people with a lot to learn from. Sharing know-how with each other is a given. Our technical co-founder, for example, has a PhD in applied mathematics. The emphasis on honest expertise is in our genes.
We primarily use SQL and Python to process big data. On the side, we have more modern technologies like Kibana or Grafana. We automate a lot of things into Django, a Python extension.
We use neural networks for inspections and have a lot of well-processed data for machine learning. We have the cloud in AWS, we do automation primarily in Jenkins, and because Tapix is a docker service, we also work with Kubernetes.
But it's not all about work and technology. We're a group of people who pursue our hobbies outside of work and don't just sit on the couch. Sports are our favourite thing, almost everyone at Dateio does sports, and this is especially true for our data team.
Some of us like cycling or the not so occasional jog, some of us play sports on a professional level. We have a colleague who is the Slovakian short triathlon champion, a climber, and a volleyball player who has participated in the world championships several times. That's why it's not a problem to park your bike or hang your sweaty jersey in our place.
Lenka Viceníková
Head of People Care
