About
We build software for people who work with texts and collections.
The problem
Mass digitisation solved acquisition. A researcher today can reach more primary material from a laptop than a scholar of the last century could reach in a career. What did not follow was a matching improvement in how that material can be searched, linked, and cited.
A scanned volume that cannot be searched is closer to an unopened box than to a book. An OCR layer with no correction pass produces confident, wrong answers. A collection with no stable identifiers cannot be cited, and so cannot really be used as evidence.
The name
In Borges’ The Library of Babel, the library contains every possible book. Somewhere in it is the true catalogue of its own holdings — and also every plausible forgery of that catalogue. Total information, zero navigability.
We took the name because the joke is on us: the modern archive has the same shape. Our interest is in the catalogue, not the stacks.
How we work
- Open where we can. Research tools that cannot be inspected cannot be trusted, and results produced with them cannot be reproduced.
- Formats over platforms. Work should outlive the software that produced it. We prefer plain text, documented schemas, and exports that do not need us.
- Narrow tools. A small program that does one part of the pipeline well composes with the rest. A platform that does everything usually does not.
- Built with the people using it. Domain knowledge lives with librarians, archivists, and editors — not with us.
Get in touch
We are in early development and open to conversations — collaborations, pilot projects, or a description of the problem you keep working around.