More in Knowledge
Datashare
Search a document collection the way an investigative desk does: point it at a folder of PDFs, emails, spreadsheets and scans, and it extracts the text — running OCR over the images — indexes every document, pulls out the people, organisations and places named in them, and lets you run saved batch searches across the lot. Built by the ICIJ for the Panama Papers, and everything stays on this machine.
Docling
Turn PDFs, Word, PowerPoint, Excel, HTML and scanned images into clean Markdown, JSON or HTML with layout, tables, reading order and OCR understood by AI models - a local document converter with a web UI and an API for RAG pipelines.
GCompris
Educational suite for children aged 2 to 10 with more than 190 activities: reading and writing, counting and arithmetic, science, geography, logic games, music and computer basics - a dashboard for teachers included.
Kiwix
The whole of Wikipedia offline, and much else besides: open ZIM archives of Wikipedia, Wiktionary, Project Gutenberg, Stack Exchange, TED and medical libraries, then read and full-text search them with no connection at all.
Open Notebook
Private alternative to NotebookLM: collect PDFs, web pages, videos and audio into notebooks, chat with your sources, write notes with AI, and turn research into podcast-style audio - with the model provider you choose, including local ones.
OpenBoard
Interactive whiteboard for teaching: write and draw over lessons, import PDFs and images, use rulers, compasses and protractors, embed web pages and interactive widgets, and record a podcast of the lesson - made for schools and universities.
