All models are wrong,

but some are useful.

— George E. P. Box, 1976

Clustering is a pipeline,

not a button.

And every stage has an LLM-era upgrade you might not be using yet.

Built with

Open source — mostly.

Hugging Face

PyTorch

scikit-learn

Anthropic Claude

pandas

NumPy

JupyterLab

★TMDB

UMAP

HDBSCAN

BBERTopic

·datamapplot

Plotly

Matplotlib

MManim

EEVoC

Thank you.

Seth Levine · Director of AI Innovation, Contentsquare

github.com/splevine/
clustering-good-bad-beautiful All notebooks, data scripts, and visualizations. Colab-ready.