Once we switched to more focused ID scan software, things started to make sense because it’s not just about reading text, it’s about understanding the document itself—like where the fields are, how they relate, and whether something looks off. In my case, I ended up using something like https://ocrstudio.ai/id-scanner/ as a reference for how these systems are supposed to work, and it helped a lot in figuring out what features actually matter, like automatic data extraction, support for different ID types, and decent accuracy even with imperfect images.