Combine output of two models to create a risk heatmap

Question

I have two models (predicting risk of large loss) model a and model b. I need to come up with a risk heatmap where I say model 1 score > x and model 2 score > y then high risk else low risk. How do I combine two different risk scores? What metrics do I need to look at? I am not sure where to begin. I just have the output of the two models - where I have the bad rate by score thresholds.

I have the bad rate for each of the models separately.

Model A 
score      bad rate
0.1-0.2    500/10000
0.2-0.3    1000/10000
0.3-0.4    1500/10000
0.4-0.5    1600/10000
     ..
     ..
 0.9-0.1    2000/10000

my y label is for every seller , whether there was a dispute or not.