Why this conversion is harder than it sounds
A Word document is a list of paragraphs. A PDF is a set of glyphs at fixed coordinates. When a document is exported to PDF the paragraphs are thrown away, and nothing in the file records that these eleven lines were one paragraph and the next line was a heading. Going back means working it out from what is left: the size of the type, the gaps between lines, where the text starts on the left.
That is what this converter does, and it is why the result is an editable document rather than a facsimile. You get the words, the reading order, the headings and the lists. What you do not get is the original page design, because the design was position, and position is exactly what has to be given up to make the text reflow again.
What comes across, and what does not
Headings are recovered from type size, so a document with a clear typographic hierarchy converts well and one set entirely in 11 point does not, because there is nothing to detect. Bulleted and numbered lists are recognised from their markers. Bold and italic come across where the PDF names its fonts in a way that says so, which most do and some do not. Words broken across a line with a hyphen are put back together.
Three things will not survive, and it is better to know before you convert. Multi-column layouts are flattened into reading order, so a two-column academic paper comes back as one column. Tables drawn with rules rather than stored as tables become paragraphs. And a scanned PDF has no text in it at all, only a picture of text, so there is nothing to extract. The tool says so rather than handing you an empty file, and the fix is to run it through OCR first.
Why it does not upload your file
Almost every free PDF to Word converter works by sending your document to a server. That is a real decision, not a technicality: the file people most often want converted is a contract, an invoice, a statement or a report, which is precisely the sort of document that should not be sitting in a stranger's upload queue with an unread retention policy behind it.
This one does the work on your own machine, using the same file interface a browser uses to show you a photo before you post it. Nothing is transmitted, so there is nothing to store, nothing to leak, and no size cap imposed by somebody else's disk. The page works offline once it has loaded.
When you want the design back, not just the words
This gives you an editable document. It does not give you a designed one, and for a lot of the reasons people convert a PDF, editing the words is the whole job and that is the end of it.
It is a different job when the PDF was a lead magnet, a guide or a report and what you actually want is to publish an updated version that looks considered. That is layout rather than conversion. Designrr imports the same PDFs and Word files and turns them into a designed ebook or flipbook, with templates, a cover designer, an automatic contents page and one-click export.