Question Customers are experiencing a reoccurrence of an error issue in the Data Extraction Scope activity while extracting document values using a specific model of MLSkill. The issue appears to occur with a specific document type that has been previously trained. Only one document is currently affected.
Issue Description + Resolution
Error Description
The error message ServiceException 400: InvalidRequest - Payload is Corrupt indicates that the transmitted data is corrupted or invalid. This confirms that the server was unable to process the request properly. The error seems to occur with a specific document type.
Potential Cause
Formatting issues or certain text elements in the PDF files may be causing conflicts with MLSkill's prediction process.
Resolution Steps
- Apply one of the PDF files to the ML Skill individually and observe the behavior.
- Test after converting the problematic PDF file to another format (e.g., PNG or TIFF image) to check if the issue persists.