Design notes

Why I built Seedex

You run into a word you don’t know. You look it up. You understand it, you’re satisfied — and then you forget it.

Seedex started as an attempt to break that cycle.

Satisfied the moment you look it up

Reading a book, reading an article, a word comes up that you don’t know. You search for it on your phone, or open a dictionary app. Once you have the meaning, it feels resolved.

The problem was that it ended there. I didn’t write anything down. And on the rare occasion I did, I’d later find myself wondering where I had put it. So I’d end up looking up the same word all over again.

The same thing happened with AI translation. A word I didn’t know would turn up in the output, I’d look it up, I’d understand it — and no record of it survived.

Scrolling back through dictionary history

At one point I tried scrolling back through my dictionary app’s history to revisit words I’d looked up. They were there, technically. But hunting for a particular word in a long list of searches was tedious.

What that history doesn’t hold is when I met the word, what I was reading, and why it caught my attention. Just the word, sitting there on its own.

I built a “Language Bank” in Notion

So I decided to build it myself, and set up a database in Notion. I called it Language Bank — a place to accumulate the words I came across.

At the very least, my words stopped scattering.

I was just satisfied with collecting them

What I realised after a while was that collecting words isn’t the same as learning them.

The collecting had become the goal in itself, and I felt done the moment I typed something in. Looking back later, all I saw was a list of strings — none of the feeling of the moment came back. Like coins dropped into a jar and never counted, accumulation on its own didn’t amount to much.

Maybe images would make them stick

That’s when it occurred to me: what if I learned words alongside images?

Words slip away partly because we try to hold them as words alone, so with a visual anchor they might stay — and if I could tie a word to a scene I’d actually seen, or to the image it brought to mind, I thought I’d have more ways back to it.

No app could match the image to my situation

I looked at existing apps too, but most are either flashcard-style, or they pair words with images that ship with the app.

Those stock images are polished and accurate. But they had nothing to do with the situation in which I met the word. A word from a sign I saw while travelling, a word from a line in the novel I was reading — tying words to images drawn from my own experience was the part existing apps made difficult.

The answer: record the encounter itself

By this point, what I wanted was clear:

  • A place to keep the words I encounter
  • The ability to attach photos I took myself
  • A way to generate an image when I don’t have a photo to hand
  • And a record of where I met the word

I couldn’t find anything that did all four, so I decided to build it. That became Seedex.

In Seedex you can attach an image to a word — a photo you took yourself, or one you have AI draw for you. Alongside it, you can keep the context in which you met the word: the passage you were reading, a URL, a page number, whatever you were thinking at the time.

The words you add grow as a tree. It makes what you’ve collected visible, and past a certain number it bears fruit — a way of not letting things end at “collected it, felt good”.

Meeting a word doesn’t finish when you’ve looked up its meaning. If you can keep the circumstances along with it, it should become far more your own. That’s where Seedex started.

← Back to the blog