How AFileFix's PDF Pipeline works

PDF Pipeline lets you chain multiple PDF operations into a single workflow. Instead of running Watermark, then downloading, then opening Compress, then downloading again, you build a chain of steps once and run it all in sequence. Drop in up to 10 PDFs, build your chain, click Run.

Each step in the chain uses the underlying tool's existing implementation — so if your chain includes Watermark (browser-side) and Compress (server-side), Watermark runs on your device and Compress runs on our server. The privacy treatment for each step matches what you'd see using that tool directly.

What steps can I chain?

Pipeline supports most AFileFix tools as steps:

Browser-side steps (your file stays on your device for these):

• Watermark
• Rotate
• Add Page Numbers
• Flatten (lock the document — no selectable text or editable fields)

Hybrid step (tries browser first, asks before falling back to server):

• Merge

Server-side steps (file uploaded over HTTPS, deleted within minutes):

• Compress
• Password Protect
• Unlock PDF
• Repair PDF

You can configure each step's parameters before running — compression level for Compress, watermark text and position for Watermark, password for Protect and Unlock, and so on.

How does Pipeline handle privacy for mixed chains?

This is the most important thing to understand about Pipeline: a mixed pipeline doesn't change individual step behavior. Each step still runs where it normally would.

A pipeline with only browser-side steps (Watermark → Rotate → Page Numbers) runs entirely on your device. Your file never reaches our server.

A pipeline that includes any server-side step (e.g., Compress) uploads your files to the server for that step. Once the server step finishes, browser-side steps in the chain resume on your device.

The privacy banner on the Pipeline page reflects this — it shows "hybrid" because the actual behavior depends on which steps you've added. For the full per-tool privacy treatment, see our Privacy page.

How do I build and run a pipeline?

1. Upload your PDFs. Up to 10 files per pipeline run, 50 MB per file, 200 MB total. Password-protected PDFs aren't accepted at upload — use Unlock PDF first, or include Unlock as a step in your chain.
2. Add steps to your chain. Click steps from the chain builder. They appear in the order you add them. You can include the same step type multiple times if your workflow needs it (e.g., Watermark twice for different positions).
3. Configure each step. For each step in your chain, set its parameters — watermark text and position, compression level, password for protect or unlock, and so on. The Pipeline UI prompts you for these before you can run.
4. Run pipeline. AFileFix processes each file through every step in order. Sequential execution: file 1 goes through all steps, then file 2, then file 3. Status indicators show progress per file and per step.
5. Download results. For a single file: just the final processed PDF. For multiple files: a ZIP archive with one PDF per source file. Each file reflects all the steps in the chain that succeeded.

What happens if a step fails on one of my files?

Pipeline is forgiving by design. If step 3 of 5 fails on file 2 of 10, the pipeline doesn't stop — it skips the remaining steps for that file and continues with the other 9 files. The results screen shows which files completed successfully and which had errors.

This matters because: PDFs vary enormously in structure. A scanned page that breaks the Compress step in one file might not exist in your other files. Stopping the whole batch over one bad file would waste the work on the good files.

Common failures:

• Compress reports "already optimized" — that file skips Compress and continues with other steps
• Unlock with wrong password fails — that file stops at Unlock and the remaining steps don't run on it
• Server timeout on a large file — that file fails its current step and the remaining steps skip
• Merge fallback to server gets declined — Pipeline runs continue, but the merge step doesn't combine your files

The results screen tells you which files failed where, so you can re-run the pipeline on just those files with adjusted parameters.

When should I use Pipeline vs running tools individually?

Use Pipeline when:

• You have a recurring workflow (e.g., monthly invoice batch: rotate scans → watermark → compress → protect)
• You're processing several files through the same set of steps
• You want to set parameters once and run them across multiple files
• You don't need to inspect intermediate results between steps

Use individual tools when:

• You only have one file and one operation
• You need to inspect or adjust the result between steps (e.g., check compression quality before applying a watermark)
• The workflow is one-off and you won't repeat it
• You want to preview a step before committing the whole chain

Pipeline rewards repetitive workflows. For one-off tasks, the chain-building overhead isn't worth it.

What types of workflows work well in Pipeline?

A few real workflows that benefit from Pipeline:

Document distribution prep — Watermark "CONFIDENTIAL" → compress → password protect. Used for sharing sensitive documents externally where you want a visible mark, smaller file size, and access control.

Scanned document cleanup — Rotate to fix orientation → repair → compress. Used when processing batches of scanned receipts, contracts, or forms before archiving.

Versioned document publishing — Add page numbers → watermark with version string → compress. Used for documentation, manuals, or reports with version tracking.

Bulk file processing — Apply the same operations to many files at once. Up to 10 files per pipeline, so very large batches need multiple runs.

Limits and gotchas

A few things worth knowing:

Sequential, not parallel. Pipeline processes files one at a time, not concurrently. A 10-file pipeline with multiple server steps can take several minutes. For most use cases this is fine; for very large batches it's a real time cost.

Maximum 10 files per run. If you have more files, run the pipeline multiple times. There's no in-app way to queue more than 10.

Maximum 200 MB total. The aggregate cap matches other AFileFix tools.

Passwords are validated at runtime. If you include Protect with a 3-character password (below the 4-character minimum), the pipeline fails at that step. Same for Unlock with a wrong password — the pipeline will tell you per file which password attempts failed.

Output filenames follow each step's own convention. If your chain ends with Compress, output is {name}_compressed.pdf. If it ends with Watermark, output is {name}_watermarked.pdf. The last step in the chain determines the filename suffix.

Frequently asked questions