I use multiple datasets, multiple algorithms and want to run the same pipeline on them, but they have a different number of tables with different names and the algorithm has a different io