To use nearly any application today, you hand custody of your data to its developer, who in turn hands it to a small number of cloud providers. You generate the data, and you are the only party in that chain without keys to it.
Most of what people dislike about the modern internet follows from that arrangement: surveillance-based business models, lock-in, data lost when a service shuts down, and recurring breaches. These are not the failures of particular companies. They are what the architecture produces.
What Isonomic changes
Isonomic separates application logic from data custody. It defines one application stack of six services: relational, graph and vector databases, a headless content service, application hosting, and language-model inference.
That stack runs the same way in three places:
- on hardware you own
- on a Cluster of machines owned by your team
- on an open network of independent operators who cannot read what they host
Software is delivered to your data, not the other way round. A developer can ship an application to any number of people and still be unable to read their data.
Read more
The full design is in the whitepaper.