Data Not Extracted from an Imported File
Quick Guide to Creating Certs When Imported Data Fails to Populate
After importing or emailing a PDF into SmartCert, all of the expected data may not be pulled from the file. The SmartCert entry is created, but some of the fields that should have populated from the imported document remain empty or incomplete.
Cause
This happens when the AI extraction step returns an unexpected response structure. In some cases this looks like a hallucination, or the model alters something it shouldn't during processing. This is a known limitation of AI-based extraction today. As the underlying models improve, we expect to see this occur less often, but it can still happen occasionally in the meantime.
Workaround
If a file has been imported and the data didn't extract properly, you do not need to download the file and re-upload it. Just follow these simple steps:
- Go to the Files section where the file lives.
- Select the file (or files) that didn't extract correctly.
- Click the ellipsis (⋯) menu next to Select.
- Choose Create SmartCerts from all files.
This reruns the SmartCert creation process on the existing file without requiring you to delete it and start over. You have the option to delete the previous incomplete SmartCert.