1
Tables & features
pick the data to cluster on
2
Derived features
AI-suggested derived features
3
Duplicates
redundant features flagged
4
Encoding
each feature → numbers
5
Dimensions (PCA)
elbow & best pick
6
Clustering (3 models)
pick each model's best k & save
7
Consensus
overlay the models → major clusters
8
Subclusters
split each cluster again → final segments
9
Visualize
final segments → features & members