I had a go at this, but it is not quite correct. The problem seems to be the B-Spline at the upper planar surface (curve B1,) which I had to approximate by eye.
I have a feeling this can be modeled, perhaps with the aid of the curves WB.
Any how, I used Part Cross-section to define A1,B1 and A2 (I had hoped to get B2 as well but no go!
I then imported fusion into PD WB and created some data features, sketches and made an additive loft.
This models the prismatic feature you were looking for but is not parametric. I think if you start in PD WB and use a spreadsheet or some expressions, this could be made parametric.
I have included a file but i think I broke it when checking for its parametricicity (is that a word?)
Hope this plants a seed!
Looks like a turbine blade or something?