This is very doable. What domain is your data? e.g. tabular, images, videos?

If it's vision, are you able to share a bit on what quality control metrics you have and whether you need to detect or classify anything in the images / video?