I will present recent progress in constructing a generic two-loop amplitude reduction algorithm within the computational framework of HELAC. Following the well-known OPP reduction approach at one loop, a two loop amplitude approach is developed. I will also discuss the differences between the 4-2ε and pure 4 dimensional reduction fitting as well as the implications on the so-called rational terms which incorporate the mismatch between 4 and D = 4-2ε dimensions.