Convert PDF to Excel without uploading your file

The architecture, in plain English

1. Your PDF
2. Your browser
3. PDF parsing / OCR
4. Manual review
5. Excel file

You choose a file (or the sample). JavaScript in this tab receives the bytes. PDF.js interprets pages on the device. If a page is a scan, a local OCR worker reads an image of that page. Candidate rows appear in the review table. When you export, ExcelJS—loaded at that moment—builds an .xlsx file and your browser downloads it. At no point is the document content submitted to a Linework API for extraction.

What stays on the device

• Document contents stay on your device during processing.
• Processing happens locally in the browser.
• There is no PDF storage on Linework.
• There is no extraction history after you close or reload the tab.
• No document is sent to an external AI service.

Preferred wording, and the accurate claim, is: your PDF is processed locally in your browser and is not uploaded to Linework for document processing.

What this does not claim

Visiting the site still uses the network. The browser fetches HTML, CSS, JavaScript, and, when OCR is needed, engine and language files. Advertising scripts may load if advertising is enabled. Hosts can log ordinary requests for those assets. Browser caches can keep app files. None of that is the same as uploading your PDF for conversion, but it also means a phrase like “nothing ever leaves your device” would be too broad.

Why teams ask for this

Supplier files sometimes include unpublished prices or customer-specific packs. Uploading them to an unknown converter creates a copy you do not control. A local workflow keeps the working copy in tab memory. You still need a safe place for the Excel file you download; that file is yours to store or delete.

How to use it

Open the converter, choose the PDF, map columns, review, export. The same limits apply: 15 MB, 50 pages, 20 OCR pages, English OCR, 3,000 candidate rows. Read privacy details, the workflow, table conversion and OCR for scans.

What a typical upload converter does instead

Many online converters ask you to send the file to a server, wait for a remote process, then download a result. That model can be convenient, and it can also create a copy of unpublished prices on someone else’s disk. Linework’s model is the opposite for document contents: the bytes you choose are read in the tab. The trade-off is that your device does the work. Large scans occupy CPU and memory here, not in a queue on a GPU farm.

Because there is no project store, there is also no “resume later” on another phone. If you close the tab, the working rows are gone. That is intentional. Download the Excel file when the review is finished. The downloaded file is an ordinary document on your disk; protect it the way you protect any price list.

Advertising, logs and caches

The public site may load an advertising script. That script is unrelated to PDF parsing. It does not receive your spreadsheet rows from Linework’s converter code. Ordinary hosting logs can still record that someone requested /style.css or /app.mjs. Browser caches can keep those program files so the next visit starts faster. None of those facts change the processing path for the PDF you open, and none of them should be summarised as “we collect nothing at all”.

If you need the same local behaviour as a packaged internal tool, you can still use this static site: there is no account wall. The supplier, catalogue and guide pages exist so you can read the method before you open a sensitive file.

Frequently asked questions