pandas: how to aggregate a subset of groupby rows into a single row?

Here's what I'd like to do given the following input:

pd.DataFrame({'cat':['A','B','C','B','C','D','C','E'], 'value':[3,6,7,7,9,8,3,1]})

cat    value
A          3
B          6
C          7
B          7
C          9
D          8
C          3
E          1

1. Group by cat and sort descending:

   df.groupby('cat').sum().sort_values('value', ascending=False)
   
   cat    sum
   C       19
   B       13
   D        8
   A        3
   E        1
   

2. Leave rows that cumulatively add up to less than 90% as is, but the remaining rows combine into a single new category 'Other':

   cat    sum
   C       19
   B       13
   Other   12
   

How do I do the last step?