Martin Kleppmann

Researcher at University of Cambridge


Automerge: A new foundation for collaboration software
distributed systems

Your rating:

Lots of software these days needs to enable collaboration between several users. We know how to build basic web apps, where all users read and update data in a shared database. But real-time collaboration, as we know it from Google Docs or Figma or Trello, is a much bigger challenge. It gets even harder if you want to allow users to continue working while offline. Automerge is an open-source library that aims to make collaboration software simple and robust. It provides a shared JSON-like data structure that several users can update at the same time, and which automatically merges all updates into a consistent view. It provides data formats for efficiently storing this data and syncing it between users. It seamlessly supports both offline work and live real-time collaboration while online. Internally, Automerge is based on Conflict-free Replicated Data Types (CRDTs). In this talk we will explore some of the interesting computer science research that makes it possible, and see how it is used in practice.


Dr. Martin Kleppmann is a researcher at the University of Cambridge, and author of the bestselling O'Reilly book Designing Data-Intensive Applications. He works on distributed systems and security, in particular collaboration software and CRDTs. Previously he was a software engineer and entrepreneur, co-founding and selling two startups, and working on large-scale data infrastructure at LinkedIn.