Evisort 'document failed to process' - the upload sits at 'processing' for hours

TL;DR

An Evisort document stuck at 'processing' for hours - or flagged 'document failed to process' - means the AI pipeline never finished it: the file is in an unsupported or encrypted format, it exceeds size or page limits, or the job stalled in the queue. Check the file against Evisort's requirements, strip password protection, and re-upload a clean copy. If a clean, supported file still stalls, the job is wedged server-side: note the document id and open a support ticket rather than re-uploading repeatedly.

The query

Evisort 'document failed to process'  -  the upload sits at 'processing' for hours

Use this when

• An upload sits at 'processing' for hours with no progress
• Evisort reports 'document failed to process'
• Other documents process fine but this one never finishes

Not for

• Processed documents with wrong extracted data - those completed, the extraction was just inaccurate
• Upload errors (network failures, rejected files) - those fail before processing starts
• Slow processing across all documents - check for a platform incident first

Steps

1. Check the file against the supported requirements

Confirm the format (PDF, Word, and the other listed types), the size limit, and the page limit for your plan. Oversized files and exotic formats fail or stall instead of being rejected cleanly at upload. Expected output: the file is inside every stated limit, or you find the limit it breaks.

2. Remove password protection and flatten the PDF

Encrypted PDFs cannot be processed - the pipeline has no password to try. Scanned image-only PDFs, portfolios, and files with heavy XFA forms also stall. Decrypt, flatten, and re-export to a plain PDF before uploading.

qpdf --decrypt --password value [DOC_PASSWORD] stalled.pdf clean.pdf

Expected output: clean.pdf opens without a prompt and is a plain, unencrypted PDF.

3. Re-upload the clean copy and watch the status

Upload the cleaned file as a new document and monitor its status through the normal stages. Do not re-upload the original repeatedly - if the file itself is the problem, retries just burn queue time. Expected output: the new upload moves through processing to complete within the normal window.

4. Escalate a clean-file stall with the document id

If a supported, clean file still sits at processing past the normal window, the job is wedged on Evisort's side. Copy the document id, note the upload time, and open a support ticket - include the file's format, size, and page count so support can reproduce the stall. Expected output: a ticket referencing the exact document, instead of a fifth identical re-upload.

Tool compatibility

Evisort document ingestion and AI processing pipeline. Companion tools: qpdf for decrypting and flattening before upload.

Variant phrasings

• Evisort document failed to process
• Evisort upload stuck at processing for hours
• Evisort document never finishes processing
• Evisort processing stalled re-upload not working

Why it happens

AI document processing is a multi-stage pipeline - parse, OCR if needed, classify, extract - and a failure or hang in any stage leaves the document at 'processing' with no partial results surfaced. Encrypted or malformed files fail the parse stage deterministically, which is why cleaning the file fixes most stalls; the remainder are genuinely wedged jobs that only the platform team can unstick.

Edge cases / pitfalls

• Very large but supported files can take much longer than small ones - learn your account's normal window before calling a stall.
• Files with hundreds of scanned pages may need OCR time counted separately from the quoted processing window.
• Deleting the stalled document does not always cancel the wedged job - check it does not resurrect as a duplicate.
• Keep the original file: support may ask for it to reproduce the stall even though you uploaded a cleaned copy.

Provenance

Resolved from the public thread: https://vectle.com/posts/pstI9838ss9Oxymv-FB6g2xw

Resolved from the public thread: https://vectle.com/posts/pstI9838ss9Oxymv-FB6g2xw

Thread query: Evisort 'document failed to process' - the upload sits at 'processing' for hours

TL;DR: An Evisort document stuck at 'processing' for hours - or flagged 'document failed to process' - means the AI pipeline never finished it: the file is in an unsupported or encrypted format, it exceeds size or page limits, or the job stalled in the queue. Check the file against Evisort's requirements, strip password protection, and re-upload a clean copy. If a clean, supported file still stalls, the job is wedged server-side: note the document id and open a support ticket rather than re-uploading repeatedly.