A pen for writing into your own memory.

Where the name comes from

Ebbio uses FSRS inside the app, but the name is a debt to the forgetting curve and to Hermann Ebbinghaus, who discovered it and started the study of forgetting.

Both the classical forgetting curve and FSRS are remarkably complete as research, and both have been remarkably little used in ordinary life. Where they were used at all, it was for vocabulary and flashcards. Ebbio exists to optimise that work for everyday use, and to make the retention and recall probability of any memory visible.

Why Ebbio was built

It started with the developer’s own problem: knowledge learned for work that could not be recalled when it was needed.

What had to be remembered was not only words. English and French vocabulary came to more than 3,000 items across the two languages, and on top of that were the steps for producing documents, the map of a new posting, and a wide range of other working knowledge. Different forms, different fields, all needed at once. After half a year of study and as much review as could be managed, more than half of what had been learned early on was gone within four months.

The first attempt was vocabulary apps. Two of them, over one or two months, before giving up. Deciding for yourself when to look at something again means forgetting it again. Neither app allowed enough freedom in labelling to organise anything that was not vocabulary.

Next came spaced-repetition apps. Two more, over three months. The app now chose the review dates, but the algorithm was old: roughly a fifth of what came up was already known. Time went into reviews that were not needed. Labelling was again too rigid for broad memorisation.

Then came FSRS: a modern memory model fitted to hundreds of millions of real reviews, and clearly the answer. But the apps that could take FSRS were few. On one of the four, installing it meant fetching an add-on and pasting code; getting to a usable state took about fifty minutes, the day-to-day feel was awkward, and it looked crude.

The largest problem remained. That app was built for vocabulary, and it did not suit procedures or maps. A modern algorithm is worth nothing if what you want to remember is not a word. And one more worry never went away: work knowledge includes confidential material, and sending it to somebody else’s server was not something to be comfortable with.

Five design decisions

1. FSRS-4.5 built in, not bolted onModern spaced repetition runs from the very first card, without the user having to think about the algorithm, choose review dates, fetch an add-on or paste any code.
2. The unit of record is not a flashcardVocabulary, procedures, maps and working knowledge all run on the same mechanism. Nothing has to be managed separately because of its shape.
3. Deep enough hierarchy for real knowledgeCollections; a tree of chapters and sections inside them; a Path such as “1-3-6” that places an item directly; and tags that cut across everything. Vocabulary, procedures and maps can live in one app as separate systems.
4. Minimal by defaultCollections, tags, goal dates, daily limits and notifications can be switched on when they are needed. The complexity of every feature being present at once is never imposed.
5. Privacy and security at the centreData is stored on the device by default and never sent to a server. PBKDF2-SHA256 (600,000 iterations) with AES-256-GCM, a 4–8 digit passcode, an arbitrary passphrase, biometrics, a randomised number pad and a second profile let you choose the level you need.

About the price

Ebbio is a one-time purchase, so that the tool is available to as many people as possible.

Memory matters a great deal in a life. Supporting it should not mean disturbing the experience with advertising. You pay once, for the maintenance and development that keeps the app usable as a memory instrument for a long time. That keeps the cost to the user low while keeping the app sustainable.

How Ebbio differs

Most tools that offer spaced repetition sit at one of two extremes: complete as research but complex enough to require real learning before use, or so committed to convenience that they are optimised only for vocabulary and language study, or built mainly for note-taking with retention as an afterthought.

Ebbio fills the gap. Research-grade mechanisms (FSRS-4.5, workload smoothing, goal-date scheduling) are implemented for any memory rather than for a narrow set of cards, while the screens and controls stay minimal and features activate only when you want them.

Preferring generality does not mean losing on the narrow ground. Even compared card for card, FSRS is a newer generation of spaced repetition than the algorithms built on the classical forgetting curve.

What Ebbio is not for

One: study built around sharing decks with other people. Ebbio is designed as a personal tool for maintaining your own memory, and group use of shared decks is not part of it.

Two: reaching the same data from several devices. As the flip side of keeping data on the device, there is no sync between devices today. A secure, selective sync (Ebbio Sync) that builds on the app’s encryption is under consideration but not implemented.

About the developer

Ebbio is made by Cyver Labs, founded by volunteers for the Ebbio project: making FSRS general-purpose for ordinary use. The project started in October 2025, programmers and designers joined, and the app was released in September 2026. Development continues.

The name Cyver Labs has two roots. One is the plain fact that the project could not exist without cyberspace: computing the state of an enormous number of memories one by one, and assembling the right review for each day, only became possible in the age of computers and networks. The other is Cy(cle) + ver(sion / veritas): the idea that memory, recalled again and again, settles as a continually updated self and as truth.