Hi, this looks just like what I am looking for! Have you thought about making the output of multiple runs comparable by supporting the export of benchmark data to e.g. CSV?