Writing and data visualization
Essays, data stories, tutorials, interviews
Writing
Opinions, essays
I occasionally write on the Substack Lookalikes and Meanders. See for example:
- 29.01.2023: “Publish-or-perish” and ChatGPT: a dangerous mix Substack
A selection of my essays published elsewhere:
- 21.02.2024: “Guest Post — There is More to Reliable Chatbots than Providing Scientific References: The Case of ScopusAI” Scholarly Kitchen Blog
Scientific writing
I am increasingly engaging in research work within scientometrics:
- 05.2025: A chapter about data curation in global repositories, published in “Recommendations for Swiss Higher Education Institutions to enhance the visibility of their research output” PDF
- 07.2023: How important are international collaborations for research in Switzerland and the UK? Blog post - a write-up of our article in “Qualität in der Wissenschaft” 01/2023 (contact me to get a copy of the original article in German)
- 2023-…: every few months we publish a report about the project “Towards Open Bibliometric Indicators” (TOBI), which I’m a lead of. See on the TOBI Website
There is also a selection of my work in physics available online:
- I am a first author of the article published in Science in 2014, which work was part of my PhD.
- My PhD Thesis is available through the Research Collection of ETH Zurich. Besides physics, the appendix discusses some of the limitations of PCA applied on noisy data, and shows dashboards for data quality monitoring in 2D experimental scans.
- In the ancient times (2003-2011) I used to be an author and editor of a leading Polish astronomy news platform, AstroNET
For the full list of publications, consult my ORCID Profile.
Tutorials
I host a blog with Python recipes and an awesome-matplotlib repository, where I collect useful snippets as well as longer-form tutorials.
Data visualization
Data stories
A selection of data stories:
- 12.2019: Mount Mid-Life Crisis Medium / Towards Data Science - uncovering insights from the Himalayan Database
- 12.2019: A story told through a heatmap Kaggle - Winner of the Kaggle 2019 Data Science Survey
- 10.2019: Ironman World Championship: Game of skill - game of luck Self-hosted - my experiment with clicky-scrolly form using a Makeover Monday dataset
- 10.2019: Top 20 CO2 polluters visualized Medium - looking for patterns behind the emissions and revenues of companies identified by The Guardian as the biggest CO2 polluters
- 04.2019: Beyond the bar plot: visualizing gender inequality in science Medium - a chart-makeover exercise. I have learned so much more about dataviz since then…
Other visuals
Examples of other data visualization work
- 07.2025: A 2-year retrospective into the open information landscape: Learnings from the TOBI project Scientific Poster, Zenodo
- 07.2023: How important are international collaborations for research in Switzerland and the UK? Swiss Year Of Scientometrics Blog
- 2010-2016: Selected scientific visuals from my PhD Thesis and MSc Thesis
Interviews
- 08.2023: I was a guest on the “Unchain your mind” podcast, in the episode Wer entscheidet, welche Werte ein System vertritt? (in German). Listen to it below or on Spotify:
- 06.2023: I was interviewed for the “Jahrbuch der Marktforschung 2023”, published by the BVM Berufsverband Deutscher Markt- und Sozialforscher (in German).
- 02.2023: I was interviewed for an Swiss Radio SRF podcast «Lieber Chatbot, wie möchtest Du für Microsoft Geld verdienen? (in German)
- 12.2022: I was interviewed for an NPR article about hallucinations in ChatGPT and their implications for academia
- 12.2022: My thread about ChatGPT was featured in a Swiss news portal Watson (in German)
- 12.2021: Melanie Geiger asked me and Christian Hopp how to bring in and support women in data science. Read the whole interview on SocietyByte (in German)
Music
I play guitar in a Swiss band ÉAN. If you like Irish folk with a jazzy twist, check our Soundcloud:
















Social media
A selection of my Twitter/X threads, mostly reporting on critical assessment of chatbots as tools for research work: