ML Engineer @ Tooploox, Data Science Instructor @ DataCamp,, Book 1:1 @

Welcome to my Medium page!


Hello there! I’m Michał, a Machine Learning Engineer. I do magic with data and statistics. On Medium, I write about all things data: machine learning algos, statistical rigor, best practices, MLOps, how-(not)-to-do-stuff, and more. You can find out more about me on my website.


I’m available for consulting. You can…

Confusing the two can be costly. See how they differ and when to use each!

Both confidence intervals and prediction intervals express uncertainty in statistical estimates. However, each pertains to uncertainty coming from a different source. Sometimes, one can calculate both for the same quantity, which leads to confusion and potentially grave mistakes in interpreting statistical models. …

Regression discontinuity designs

Image by the author.

The word “because” tends to get overused significantly. We often don’t realize the strength of its meaning. “Because” implies causality — the relationship between cause and effect, which takes some statistical virtuosity to establish.

This is the third article in a series in which I discuss four statistical tools that…

The golden standard of randomized experiments

Image by the author.

Being able to establish causality is powerful. It gives you the right to use the word “because” in a conversation. Our sales increased because we have changed the website layout. The crime rate dropped because of the new preventive policy that has been introduced. Pinpointing causal relations correctly is crucial…

Artificial Intelligence

Five practical lessons and warnings for data scientists

Photo by Viktor Forgacs on Unsplash

The sudden hit of the COVID-19 pandemic found the doctors and the hospitals completely unprepared. Too little was known about the new virus and too many patients were queueing at the door to diagnose and triage them correctly and quickly enough. AI to the rescue!

Machine learning models for diagnosing…


What to watch out for while maintaining ML systems

Some of the problems described in “Hidden Technical Debt in Machine Learning Systems”, image by the author.

It is not easy to develop and deploy machine learning models, and even less so to integrate them with the surrounding data pipelines to build large-scale ML systems. The hardest part, however, comes later, when the entire system has been tested, deployed, and is up and running. For deployment is…


Tackling data drift and concept drift in production ML systems

Photo by Danny Sleeuwenhoek on Unsplash

You have collected and cleaned your data, experimented with various machine learning models and data preprocessing variants and fine-tuned your model’s hyperparameters to finally come up with a solution good enough for your problem. Then, you’ve built a robust, automatic data pipeline, wrote an API for the model, put it…

Michał Oleszak

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store