Martin Muller

Hi! I'm currently building dreamlook.ai, a highly optimized service to train customized AI image generation models. Before, I worked on Cedille.ai, the largest openly available French language model.

I started my studies at ETH Zurich, where I completed a BSc in Biochemistry. After long days in the laboratory, I had to admit that wet lab work wasn't for me and that I much more enjoyed the data side of things. I then completed a MSc in Compuational Biology & Bioinformatics at the same institution.

In my Ph.D. at EPFL, I have worked on applied machine learning and opinion mining in the context of Digital Health and built Crowdbreaks, a Machine Learning system that combines crowdsourcing and Natural Language Processing for monitoring health trends in real-time. COVID-Twitter-BERT, a language model finetuned on Twitter data collected through Crowdbreaks, has since been downloaded over 1.2M times from the HuggingFace hub.

I'm a big proponent of Open Science and improving reproducibility in scientific work, a topic on which I have organized multiple workshops.

I easily obsess over food, beer brewing 🍺 and mushroom hunting 🍄. I currently live in Princeton, NJ.