Hiring data scientists: where they are and what they build

How we recognise a data scientist on GitHub

We count someone as a data scientist when their public work is analysis and modelling: notebooks, statistics, forecasting, feature engineering and classical machine learning.

In practice, their repositories, descriptions or bio carry at least two such signals: pandas, NumPy, scikit-learn, XGBoost or LightGBM, Jupyter, time series, visualisation with matplotlib, Plotly or Tableau. Building the pipelines that move data is classified as data engineering, and deep learning research as AI research.

The stack inside the role

Python dominates, so the useful question is what sits around it: the modelling and visualisation libraries their repositories rely on.

Which countries have the most data scientists?

Data science teams exist in every economy, from banks to retail. Read the count for volume, and the index for the countries where analysis work is over-represented among developers.

Min. 1,000 indexed developers and 30 data scientists per country; counts rounded down. Method: by-country study.

The United Kingdom stands out: 2.2% of its indexed developers, 2x the index average.

Cities

Where the role overlaps with others

A profile can carry two roles. 32% of data scientists also qualify for AI research: a spec asking for both describes a real, if narrower, profile.

How reachable are they?

Much of a data scientist's work lives in private notebooks and dashboards. A thin public profile is normal here; a public analysis with its reasoning written out is worth more than a long list of repositories.

Who employs data scientists today?

What they build themselves

For the language side of the same search, see the guides to hiring Python developers. To write the query by hand first, use the GitHub user search builder.

Frequently asked questions