What are the four thresholds of IDP?

IDP will go through four stages starting with pre-processing to improve the quality of input documents to classification. Next is extraction using OCR to ensure files are machine-readable. ML extracts relevant data. After that is post-processing where the documents are validated via pre-defined business rules. In addition, an HITL framework can validate data further with its AI subset learning from human interaction.