Atomic Data

Semantic & Data Protocol

Modular specification for sharing, modifying, and modeling graph data combining the ease of use of JSON, the connectivity of RDF (linked data), and the reliability of type-safety. A strict subset of RDF with built-in schema validation via Atomic Schema (Classes, Properties, Datatypes). Uses links (URLs) to connect pieces of data, making it easier to connect datasets even across separate machines. Extended modules cover Commits (signed state changes with version history), Agents (authentication via public/private keys), Collections (querying/filtering/sorting), Hierarchies (authorization), WebSockets (real-time updates), and file management

Empresa

Detalhes

Licença MIT (specification and AtomicServer); also GPL-3.0+ for atomic-server Rust crate
Status de Dev 🧪 Alpha
Detalhe do Status de Dev Alpha (no projects have reached v1.0; breaking changes expected; AtomicServer described as 'lightweight, yet powerful CMS / Graph Database'; status: alpha on GitHub)
Proprietário Joep Meindertsma (creator, CEO of Ontola.io); Paul Opass (front-end contributor); Ontola.io (Netherlands-based software development firm); W3C Atomic Data Community Group; also founder of PauseAI
País Netherlands (Utrecht)
Ano de Início 2020
Stack Rust (AtomicServer backend: actix-web for HTTP, sled for storage, tantivy for full-text search); TypeScript/React (atomic-data-browser frontend); Svelte (@tomic/svelte library); JavaScript (@tomic/lib, @tomic/react, @tomic/svelte SDKs); CLI (atomic-cli via cargo)
Financiamento NGI Zero (3 grants: NGI DAPSI for Atomic Data core, NLnet for Solid Search, NLnet for AtomicTables)
Última Investigação 15 de jan. de 2026

Capacidades

Local-first storage Interoperable formats Self-hostable

Semantic & Data Protocol Atributos

Origins Linked Data / Semantic Web improvement (inspired by Tim Berners-Lee; takes best of Linked Data but improves with JSON compatibility, type safety, and clear consistent protocol; addresses RDF's technical complexity that limits developer adoption)
Database Sled embedded database (file-based key-value store in AtomicServer); graph database model; full-text search via tantivy (<3ms responses); 8MB download, no runtime dependencies; <1ms median response time
Query Language Atomic Paths (traverse graphs via URL path segments) + Atomic Collections (querying, filtering, sorting, pagination) + full-text search with fuzzy search and operators; SPARQL-compatible via RDF interoperability
Data Formats JSON-AD (Atomic Data JSON serialization — JSON where each key is a URL of an Atomic Property; ensures type-safety via Properties with datatypes, shortnames, and descriptions) + RDF (strict subset; can convert to/from RDF formats)
Collaborative Live Editing Supported (real-time sync via WebSockets in AtomicServer; Commits track all state changes with cryptographic signatures for audit trail and version history)
Rich Text Editing Yes (AtomicServer GUI includes documents, tables, and chat with rich editing; AI assistant with MCP support)
Mobile Support Planned (atomic-data-browser designed as web app, responsive for mobile; no dedicated native mobile app)
Web Support Yes (atomic-data-browser bundled with AtomicServer; features group chat, tables, documents; demo at atomicdata.dev; React and Svelte SDKs for building web apps)
Native Apps AtomicServer (desktop/server; runs via Docker or binary; 8MB lightweight); atomic-cli command line tool
Terms Self-hosting (run own AtomicServer via Docker or binary; atomicdata.dev as public demo instance)
Funds ~€100k (NGI DAPSI + 2 NLnet grants via NGI Zero programme)
Based On RDF / Linked Data (strict subset of RDF; JSON-AD as default serialization; Atomic Schema for type-safe classes/properties/datatypes; Atomic Commits for signed state changes inspired by Git; comparison pages for RDF, Solid, JSON, IPFS, SQL, and graph databases in documentation)
Semantic Web Compatibility Full RDF compatibility (strict subset of RDF; can convert to/from RDF; designed to be easier to use than raw RDF while maintaining linked data connectivity; W3C Atomic Data Community Group)
Protocol Maturity / Standardization Proprietary Open Standard (published specification at docs.atomicdata.dev; developed by Ontola/Joep Meindertsma)
Vocabulary / Ontology Type General-purpose / Graph database and linked data; Atomic Schema provides type-safe ontology with Classes, Properties, and Datatypes; built-in Ontology Editor in AtomicServer for creating and sharing custom data models
Usage Pattern Standalone (AtomicServer as graph database + headless CMS + all-in-one workspace) + embedded (atomic-lib Rust library, @tomic/lib JavaScript library for building apps); use cases include headless CMS, personal data store, e-commerce, surveys, verifiable credentials, data catalogs, education, food labels