Data Scientist

We’re looking for a Data Scientist to join our research team – someone who can systematically extract value from the data we collect every day.

Our deployed sensors generate a huge amount of telemetry, and so far we only use a fraction of it. We’re looking for someone who isn’t afraid to dig into data – from spotting behavioral patterns to building detection models. Your first concrete task will be classifying medical devices based on how they behave on the network; after that, you might get into probabilistic detection of dictionary attacks. And six months from now it could be a completely different topic – customer segmentation, say. Don’t worry about the domain – with us it’s mainly about working with data.

What do we do?

  • Analyzing and mining telemetry from production deployments – cleaning, aggregating, and looking for behavioral patterns and anomalies
  • Classifying devices on the network based on how they communicate and behave (our first use case)
  • Designing and implementing detections over network data – for example, detecting dictionary (brute-force) attacks with a probabilistic approach instead of fixed thresholds
  • Doing feature engineering, training and evaluating ML models (precision/​recall, false positives), and taking them all the way to production
  • Bringing AI tools and LLMs into the analytics workflow – we care about deployable results, not research for its own sake
  • Working with the development team so results actually make it into the product

How our team works

  • The research team is part of R&D and works across the whole product
  • You get data and a question – and you come back with an answer and a proposed solution
  • Team lead Martin helps you tune the direction; otherwise you have plenty of room to work independently
  • We’ll teach you the networking and security domain – we expect the mathematical thinking and data skills from you

Who we’re looking for

  • A hands-on analyst who’d rather build a working detection on real data than endlessly fine-tune an academic model
  • You have a strong foundation in statistics and math – probability, linear algebra, hypothesis testing
  • You’re comfortable with machine learning and Python in practice
  • You’re not afraid of large, real-world, not-entirely-clean data – SQL, time series
  • You can pick the right method and take an analysis all the way to a working result (the “good enough, let’s ship it” mindset)
  • You’re curious and enjoy digging into data
  • Czech is a plus, but you’ll be fine with English too

Nice to have (but not required)

  • Knowledge of computer networks (TCP/IP, protocols, NetFlow/​IPFIX)
  • A grasp of cybersecurity (attack types, threat detection)
  • Experience with anomaly detection or security telemetry

What you can look forward to

  • 5 weeks of vacation + 4 sick days – when you’re not feeling great, no guilt trips.
  • Flexible working hours and home office – most of the team is in Brno, and we meet in the office about three times a week.
  • Annual bonus – tied to company performance.
  • Education budget – for courses, books, and conferences that matter to you.
  • Multisport card contribution – to stretch your back after sitting all day.
  • Meal allowance, company laptop, and phone plan.
  • English lessons – tailored to your level, individually or in groups.
  • Team events – sometimes board games, sometimes BBQ, sometimes just a beer trip.
  • Great coffee – and we mean it. Not a corporate cliché

We are looking for a colleague for the position of

Data Scientist

Full-time position
Location: JIC, Brno-Medlánky

What’s next?

Send us your CV or a link to your GitHub/​LinkedIn – and feel free to add a few words about yourself. The first round is with HR. Between rounds you’ll get a small hands-on case study using real data – it takes about an hour or two, and it’s a nice way to see the kind of data you’d be working with here. The second, technical round then takes place with Martin at our office in Brno. And who knows – maybe we’ll be shaking hands soon.

One file only.
10 MB limit.
Allowed types: pdf.