UntrackedGuides › Keeping documents private

Keeping documents private online — the local-first guide

Every week, millions of people hand pay-slips, contracts, passport scans and invoices to free online tools — because the tool asked politely and the task was small. This guide is about the moment of upload: what actually happens, what the realistic risks are, and the design that removes the question instead of managing it.

The idea behind Untracked · 7-minute read · no account needed to verify anything here

On this page

  1. The invisible default
  2. What actually happens on upload — the lifecycle
  3. The documents that shouldn't enter that chain
  4. Local-first: the design that removes the question
  5. A practical checklist for sensitive documents

The invisible default

The default architecture of 'online tools' is: your file travels to a server, software there does the work, a result travels back. It's invisible — a drag-and-drop feels as local as saving a file — but it's a handoff. Your document is now: transmitted over the network, stored (at least transiently) on machines you don't control, processed under a privacy policy you almost certainly didn't read, in a jurisdiction you may not share.

Most of the time, nothing goes wrong. The point isn't that these companies are malicious — it's that the design makes your caution irrelevant. You can be perfectly careful and still have your file's fate decided by someone else's retention policy.

What actually happens on upload — the lifecycle

StageWhat typically happensYour control
TransitHTTPS-encrypted upload to the vendor's cloudEncrypted in transit — but the endpoint is theirs
ProcessingSoftware on their servers reads the full fileNone — you trust their code
Storage'Deleted after 1 hour' or '2 weeks' per the privacy policy — backups and logs exceptedTrust + a deletion policy you can't audit
Third partiesCDNs, cloud platforms, sub-processors — each layered in the policyEffectively none

None of this makes upload-based tools evil. It makes them a chain of trust — and chains are only as strong as the trust you'd extend to every link, forever, for every file.

The documents that shouldn't enter that chain

Ask one question: if this file leaked, what's the damage?

Local-first: the design that removes the question

Local-first tools flip the architecture: the software comes to the file, instead of the file going to the software. Two implementations:

The verification standard: a tool earns trust when you can watch it earn it. Open your browser's network inspector on any tool on this site — after the page loads, it makes no requests. Local-first isn't a promise; it's an architecture you can inspect.

A practical checklist for sensitive documents

  1. Prefer tools with no upload. If a tool works with Wi-Fi off, your file physically can't leave.
  2. If you must upload, use the minimum viable file. Crop to the needed page; redact beyond the needed field.
  3. Read the deletion promise — 'deleted immediately after processing' beats 'deleted after 14 days' beats silence.
  4. Strip metadata before sharing photos — see the photo metadata guide for what's hiding in ordinary pictures.
  5. Own the tools you use weekly. A one-time purchase beats a hundred small handoffs — that's the whole Untracked model: 43 tools, buy once, own forever.

Own your tools — the Untracked model

Every free tool on this site is also a one-time download: the same tool as a single HTML file, no daily limits, offline forever, on any computer you own. Pay once. No subscription, no account, no cloud — ever.

Browse all 43 tools →

Questions people ask

Is 'local-first' just a marketing term?

It describes a checkable architecture, not a vibe: a local-first tool works with your internet disconnected, because the computation happens on your machine. Test any tool on this site — load it, go offline, keep working.

Are browser tools as safe as desktop apps?

For file processing, yes — the page contains the whole program, sandboxed by the browser. The browser model is arguably safer than installing random executables, and the single-file downloads here are the same code as the web tools.

How do local-first tools make money if everything's free?

The free pages are limited (3 files a day); the same tools as offline single-file downloads are €5–9 once. You pay when the tool has proven itself — no subscription, no data monetisation, no upsell engine.

What about PDF tools that need heavy processing?

Rendering PDF pages for display and reordering is well within browser capability — which is why the merger here works offline. Genuinely heavy jobs (OCR on huge scans) are the honest edge case for in-browser tools; that's where desktop-class local software wins.

Related guides and tools

Free PDF Merger (no upload)
Merge, split, reorder PDFs entirely in your browser.
iLovePDF, without the upload
The comparison: same workflow, minus the server.
Free EXIF / GPS Stripper
Strip photo metadata before anything leaves.
Free Contact List Cleaner
De-duplicate contacts without uploading them.

All free tools →  ·  Untracked — buy-once software

Untracked builds single-file tools: no installation, no account, no cloud, no subscription — and no tracking on this page. Every claim here is about tools you can open and verify yourself.