About me

TLDR: I am a data scientist. I like to work at the intersection of science and engineering. I stay up to date on trends in machine learning to bring them into production.

Long story:

The term “data scientist” has become somewhat ambiguous. But I still use the term, and here is why.

Like many data scientists, I did academic research before working in the industry. My background is in cognitive neuroscience. During my PhD, I studied the human visual cortex (the part of your brain that gives you the ability to see), using functional magnetic resonance imaging. I conducted experiments, but spent most of my time analyzing data. Because I was using a novel measurement technique, I wrote analysis tools myself, and like many (data) scientists, I wrote them in Python.

After my PhD, I worked in start-ups and small companies, so my responsibilities were pretty broad. Among others, I worked on:

• Machine learning / MLOps: Identifying use cases for ML models, preparing datasets, training / finetuning models, deploying models (both computer vision and language models)

• Implementing vector search

• Refactoring & re-deploying legacy production ML models

• Data engineering: Building ETL / ELT pipelines, supporting data analysts

• Supporting decision on data science topics

• Statistical modeling / inference (admittedly that's something I'm doing less and less these days)

In other words, I worked on a broad set of topics over the past years. Therefore, “data scientist” is still a fitting job title for me. I've recently heard the term “full stack data scientist”; I like it, but isn’t it a bit redundant?