One engineer, one bench, one cipher

// WHY I BUILT THIS

“Build encryption small enough for the devices nobody can afford to replace.”

In the summer of 2025 I was contacted about building a digital notary: smart contracts, on a blockchain. The team wanted quantum-resistant crypto because of harvest-now-decrypt-later. Then they ran the numbers on microtransactions. Someone slowly buying $5 of tokenized gold can’t pay a $3 fee to do it. So they dropped post-quantum and stayed with the standard stuff.

I didn’t drop it. I kept asking a smaller question: if a key exchange happens once, what’s the cheapest way to move everything after it? No padding. No extra packets. Nothing to fragment.

The project fell apart. The cipher didn’t. I’ve worked on it every day and every night since, and I’ve rarely missed one. That’s more than a year now. Nobody else funded it and nobody else wrote any of it. It’s mine, and I’ve kept it that way. The design started from a geometric question, not from an existing cipher family: how do you build a state that never revisits itself?

The place it belongs turned out to be somewhere else. I’ve worked in manufacturing plants. PLCs with almost no memory. SCADA gear nobody has updated since the install. Sensors on coin cells that are supposed to last for years in heat that kills most electronics. Then look around your own house: gas sensors, water shut-offs, power cut-offs, remote disconnects. If someone takes over a city’s controls, that isn’t a data breach. That’s gas and water and power.

Those devices can’t hold a big cipher. They can’t afford a message that arrives in four packets when one gets lost. And nobody is going to rip them out and replace them to fix it.

The same constraint shows up in military hardware, where size, weight, and power decide what you can carry. Less crypto overhead means a smaller board, a smaller battery, and fewer things to replace.