Why I stopped trusting "free online PDF/Image tools" (and built my own instead)
A few months ago I needed to quickly compress a video and merge a couple of PDFs for a client
deliverable. Like most people, I googled "free pdf merger online" and used whatever came up first.
It worked fine. But afterward I got curious and actually read the privacy policy of that tool.
Buried in there: files are stored on their servers for "processing purposes" for an unspecified
retention period, and shared with "third-party service providers."
That got me digging into a bunch of other popular free online tools — PDF mergers, image
compressors, video converters, JSON formatters. The pattern repeats almost everywhere:
- Your file gets uploaded to their server to be processed
- Most don't clearly say how long they retain it
- A lot of "free" tools are free specifically because your file/data has value to someone
- Even reputable tools sometimes log metadata even if they don't keep the file itself
For most files this is a non-issue — nobody cares about your random meme compression. But
I started thinking about the other cases: contracts, ID scans, financial documents, internal
company files, client deliverables under NDA. People run these through random tools all the
time without a second thought.
That itch turned into a genuine "why doesn't this exist properly" moment, and eventually into
building KuberAgent — a set of 129 browser-based tools (PDF, image, video, audio, dev utilities,
finance calculators) that run entirely client-side using WebAssembly. No file ever leaves your
device — there's no server for it to go to in the first place, so there's nothing to log,
retain, or leak.
Curious what this community thinks though — genuinely, not just self-promo:
- Do you actually check privacy policies before using "free online tools," or just use
whatever's fastest? - For indie/solo devs — is "privacy-first, no server" a real differentiator people pay
attention to, or is convenience always going to win regardless? - Anyone else gone down this same rabbit hole and ended up building something because of it?
Happy to share more about the technical side (WASM performance tradeoffs, what's actually
possible to do fully client-side vs what genuinely needs a server) if people are interested.
try this :- kuberagent.com
ngl I never trusted those tools online anyway. especially the email walled ones.
Yeah that distrust is exactly why I ended up building KuberAgent where no signup, no upload, runs fully in-browser. "Free" tools where your data is the actual product got old fast.
now that ai makes these WASM based apps super easy to build I have been leaning into creating them on my site too. The idea is that they bring people to eventually see my product. Cost is nothing as it's just someone visiting the website so no matter how many people use it for free doesn't affect me negatively and adds some good to the internet. I am all for it.
Out of curiosity for yourself is it to eventually drive traffic to another site you have? Or simply putting it online because it's possible now. Cool to see you have things like background removal. I didn't realise those models have become small enough for in browser work.
Not really a funnel buddy , KuberAgent is the product itself, not a bridge to something else.
Might add a small side link someday, but that's not the point of the tools.
40 days in and honestly surprised , 4k+ unique visitors, 10k+ page hits so far, mostly organic + some community posting. Your point on near-zero marginal cost is exactly why this model
works