I built 25+ free browser tools with zero backend — here's what I learned
Every "free" online tool I used wanted something: my email, my files uploaded to their server, or my data for ads. So I built Tulyro — a collection of 25+ free tools (calculators, text utilities, developer helpers, generators) where everything runs 100% in your browser.
Zero backend, on purpose
There is no server. No database. No accounts, no login, no saved history. Each tool is client-side JavaScript that executes on your device — the EMI you calculate, the JSON you format, the image you resize never leaves your tab. That is not a privacy policy promise; it is architecture. There is simply nowhere for your data to go.
The site is a static build hosted on a CDN. Hosting costs effectively nothing, which is why everything can stay free with no ads and no "premium unlock" dark patterns.
What I learned
1. Privacy is a feature, not a footnote. "No signup" turned out to be the strongest value proposition — stronger than any individual tool.
2. SEO for tool sites is a different game. Every tool page is programmatic, so I paired each with a genuinely useful guide (how EMI prepayment works, the 75% college attendance rule, SIP vs lumpsum). The guides teach; the tools let you act. Internal linking between them does the rest.
3. Trust is the real currency. New domains get flagged by automated scanners (my 7-day-old domain got a trust score of 0 on ScamAdviser — automatically, like every new domain). Real trust signals — an about page with a named human, a contact email, valid SSL, honest copy with no fake claims — are what actually matter.
4. Speed is a feature. No server round-trips means instant results. Users feel it.
Try it
Tulyro is live at https://tulyro.online — free, no signup, no watermark. If a tool you need is missing, tell me in the comments and I'll build it.
Built by Sameer Ahmad. Everything you type stays in your browser.
Top comments (1)
"There is simply nowhere for your data to go" is a strong claim, and a Content-Security-Policy can make it easier to check, though not prove it. connect-src 'none' restricts only the script APIs MDN lists for it: fetch(), XMLHttpRequest, WebSocket, EventSource, sendBeacon and the ping attribute. It can be set in a meta tag, which suits a static build on a CDN.
It does not cover other ways a page can send data out, such as image loads (img-src, which falls back to default-src), form submissions (form-action, which does not fall back to default-src) or navigation. So connect-src alone does not show that no data leaves the browser. To back the claim you would set default-src 'none' and allow only what the site needs, add form-action, and put the meta tag first in the head, since a meta-delivered policy only applies to content after it (W3C CSP3).
I only read the docs (MDN and the W3C CSP3 spec) and did not inspect the site's headers or code.