I find it easy to extract details from both structured and unstructured documents using UiPath Document Understanding. The ML extractor is the best part; we can train our model and then use it in our process. The initial setup was good, and the documentation helped a lot. Review collected by and hosted on G2.com.

It is difficult to use with handwritten documents as we cannot annotate those. The LLM or ML engine working on the backend should be improved, so that it can recognize the fields without much annotation correctly. Review collected by and hosted on G2.com.