Built for the people
who write about reading.

I'm a programmer by trade — twenty-five years of it, mostly building web and mobile products for clients ranging from The Coca-Cola Company to the Chamber of Commerce of Serbia. These days I also advise companies and individuals on how to work better with AI. But reading is the other thing I've done all my life, and I've never made peace with screens. I read in Serbian and in English; I hold a Cambridge C2 certificate, the highest level Cambridge awards, and I still hit words I don't know.

It happened on a cold winter Sunday afternoon, in my bedroom in Vienna — warm cup of tea beside me, the page of a Hugh Howey novel open on my lap. Wool, the first of the Siloseries. A word I didn't know. The spell broke. I reached for my phone, unlocked it, opened a translation app, framed the page in the camera, hunted for the word inside the OCR result, parsed the answer, and tried to find my place again. Fifteen, twenty seconds. Every time. The flow was gone.

I knew, as a builder, that this didn't have to take fifteen seconds. Two or three was technically possible. So I set out to build the version that did — and brought two colleagues with me, Milan Anđelković and Jelena Ristić Opačić.

We named the app Lupa — magnifying glassin Serbian — and let the name become the brief. The live camera view bows outward like a real lens. The word-selection bubble springs into place like a drop of water and wobbles softly as you tilt the phone. The front camera can render itself as a faint reflection inside the glass — a quiet detail for the reader who notices. We carved away every step that wasn't tap-the-word, see-the-meaning.

By the first build, I was reading differently. My colleagues were too. We didn't know whether anyone else would feel the shift until beta testers told us — and we were surprised by who. Casual readers, yes; but also professional translators. One of them — a literary translator we'll call R.T. — told us Lupa gave them in seconds what their toolchains needed minutes for. That changed the product. Pronunciation in a chosen voice, full lookup-history export to JSON or CSV, the Glossary — every one of these shipped because a beta tester asked for it. The apparatus a working translator actually needs, built in dialogue with the people doing the work.

Now we're widening the lens. More languages, more reading surfaces — paper, Kindle, screens. The goal is the same: take the friction out of reading so a curious mind can stay curious, and the spell of the page never breaks.

— Tihomir Opačić, Founder, Lupa Vision · Vienna