Neat! I built something like this long ago (https://clustercompute.com/) to digitize a very large number of CDs for a Dutch broadcasting corporation. If I were to tackle that project today this is probably how I would do it.