from skrub import GapEncoder
gap = GapEncoder().fit(df["employee_position_title"])
encoded_labels = gap.transform(
    df["employee_position_title"].head()
)
plt.imshow(encoded_labels)