Data scientist, interaction designer, machine learning engineer. Analyst & communicator.
Our ability to construct is outpacing our ability to understand. I want to help us excel at both.
Consider the practice of feature engineering
Feature engineering is one of the key skills a data scientist has to learn to use data tools effectively. It builds a basic intimacy with the data that is lost in its final product. Modern machine learning tools are being used to deprecate this crucial practice. Engineers are taught to dump uncleaned, unexamined raw data into machine learning models, which are able to build feature representations nonetheless. OpenAI and others have not provided us with open-source, accessible language models. It’s our job now to provide our own.
The ends don’t justify the means
Or, more exactly, the means are the ends. The secondary impacts of our inventions are, in fact, just as primary as the primary impacts. My belief in openness, freedom, or any other ideal is conditional on the reality that those ideals make people’s lives materially better.
My CV
I most recently worked as a Data Engineer at Optimove, where I build and configure pipelines that connect client databases to our data visualization and interaction product. I help people connect to their own data to better reach their customers and maximize their value.