Building Bluebird: 230+ free tools, one small team

Bluebird started on a Sunday evening in late 2024 with someone else's problem.

My cousin was trying to email five photos from her wedding to her mother-in-law. Each photo was around 8 MB straight off her phone, the mail bounced at the 25 MB attachment limit, and she opened the first result on Google for "compress photo online". She uploaded all five wedding photos to a site she'd never heard of. Halfway through, a modal popped up: "Register for a free account to download your files." She closed it. The photos were still on their server.

She called me. I fixed it in about ten minutes with a tiny script on my laptop. But I couldn't stop thinking about the five photos still sitting on that stranger's disk somewhere in Frankfurt, and about the millions of people who don't have a cousin they can call.

That's the moment Bluebird existed as an idea. Not as a company, not as a product — just as a very specific irritation: this should be a five-second job that costs zero and asks for nothing, and the entire internet is set up to make it a five-minute job that costs your privacy.

The first version of Bluebird was three tools in a single-page site, built in a weekend. Image Compressor, Image Resizer, and Format Converter. I picked those because they were the tools I'd wanted for my cousin, and because I already knew the libraries (browser-image-compression and the native canvas API) well enough to wire them up without reading docs.

I hosted it on a free Cloudflare Pages account, gave the tab a hand-drawn bird logo, and sent it to about ten friends with a note: "try this, tell me what breaks".

Two things broke. The compressor produced a blurry mess above 4 MB inputs because I hadn't set a sane max dimension. And on my mum's ancient laptop, the whole page took eight seconds to load because I'd shipped a 900 KB icon font. Both were fixed by Monday morning. What didn't break was the reception: everyone who tried it said some version of "oh, this is nice, no ads?". Nine of the ten told at least one other person.

Within a month, we'd added another six tools and started getting emails from strangers. That was the point where I sat down with my co-founder and wrote three sentences on a whiteboard that have been the entire product strategy ever since.

**Free forever.** No subscriptions. No trials. No Pro tier. If a tool costs money to run on our side, we don't ship it. This is a hard rule and we've turned down real feature ideas because they would have broken it (server-side OCR, for one; a shared cloud clipboard, for another).

**Nothing uploads.** Every tool runs inside your browser tab. Your files never touch a Bluebird server, because there is no Bluebird server for user files. This limits what we can build — some things are genuinely easier on a server — but it means we can look every user in the eye and mean it when we say "your file didn't go anywhere".

**Readable by anyone.** If a nine-year-old can't figure out a tool without help, it isn't shipped. We test this literally, with actual kids, whenever we launch a new tool. It's the reason our tool pages have three big steps instead of ten small options, and the reason we banned words like "metadata", "compress ratio", and "MIME type" from the visible UI.

The catalogue grew in bursts. The pattern was almost always the same: someone would email us asking whether we had X ("a QR code generator that doesn't track me", "a way to merge two PDFs on my iPad", "a signature maker for a lease agreement"). If X could be built inside these three rules, and if we could imagine ten other people wanting it, we built it.

We're now up to eleven categories: image tools, PDF tools, YouTube helpers, calculators, text tools, developer tools, colour and design, converters, generators, security, and a growing AI section that only ships models small enough to run in a browser tab (a hard cap of about 20 MB — big enough to be useful, small enough to be respectful of your data plan).

The most-used tools are the same ones you'd expect: image compressor, PDF merge, QR code, JSON formatter. But the site's personality lives in the long tail. There's a Zakat calculator used by tens of thousands of families in Ramadan. There's a Pig Latin translator that spikes every August when kids get homework. There's a reaction-time test that half the team uses as an unofficial "am I still awake?" button.

We get asked this at least once a week. The honest answer is: the site is cheap to run and we've been patient about growing it.

Bluebird has no user database, no logins, no file storage, and no server-side compute for anything a visitor does. The site itself is static HTML, CSS and JavaScript served from a global CDN. The bandwidth cost per user is close to the cost of a single fresh page load — a few cents per thousand active users per month.

We do not run ads today. If we ever do, they'll be a single ethical placement (Carbon-style: one small text ad, no scripts, no tracking), and we'll say so on the site the day it goes live. We'll also add a "buy me a coffee" button for anyone who wants to chip in a few rupees. That's the entire monetization roadmap. No Pro tier. No subscriptions. Ever.

Three things that surprised us, in case they're useful to anyone else building for a broad audience.

First: people over 60 are a bigger user segment than we expected. They arrive on the site through the same searches everyone else does ("how to merge two PDF files"), and they stay because nothing on the page is trying to trick them. We've optimised the type sizes and contrast specifically for a 65-year-old's eyes, and everyone benefits.

Second: doing everything in the browser is not just a privacy story. It's a reliability story. Our uptime is basically 100% because there's almost nothing to fail — no server to crash, no queue to back up, no database to grow slow. When someone in a coffee shop with flaky wifi merges a PDF, it just works, because the wifi only had to load the tool once.

Third: writing plain-English UI is much harder than writing technical UI. It takes three rewrites and a testing pass with a non-technical friend to turn "Convert PNG to JPEG with lossy compression at quality 85" into "Change your image into a smaller file that any phone can open". Every one of those rewrites was worth it.

We are three people. One of us writes most of the code, one of us designs and writes the copy, and one of us does everything neither of us wants to do. We work from Bangalore, Pune and (currently) a co-working space in Lisbon. There is no VC money, no incubator, no advisor list. We ship on evenings and weekends, and increasingly during the day.

That's why we're careful about what we add. Every new tool is a new page to maintain, a new set of edge cases to test, a new bug report to eventually receive. We'd rather have 230 tools that all feel considered than 1,000 tools that half-work.

The next twelve months, in rough order of likelihood: more browser-based AI (small language models for text tools; whisper-tiny for on-device transcription; U²-Net for background removal); more calculators for real life (loan restructuring, EMI vs. rent, income tax across regimes); a proper contributors page where developers who ship a tool that fits our promise get their name and profile on the site permanently; and a light educational series — blog posts like this one, plus short videos — explaining how the tools actually work.

We'll also, at some point, sit down and prune. There are probably ten tools on Bluebird right now that don't earn their keep and should be quietly retired. Fewer, better tools is a direction we're comfortable walking in.

Three ways. Use the tools, tell one person who'd like them, and let us know if any of them break. That's genuinely the most useful thing.

If you build things: the Submit a Tool page explains what we accept and what we don't. We're picky, and we say no politely to most submissions, but a well-scoped tool that fits the three promises can be live within a week of you sending it.

If you don't build things: an email saying "this tool saved me half an hour today" lands in our inbox and makes our week. We read every one.

Bluebird is a small site with a simple mission: put good software in front of people, ask nothing from them, and get out of the way. Thanks for reading this far.