About
Why EVOSCA exists
Because on Oracle and PostgreSQL, the interesting schema work happens inside the database — and most tooling assumes it happens in a file.
Teams running serious database estates tend to end up with the same arrangement: developers share a handful of databases set aside as development environments — sometimes even the same schema — someone reconciles what they did, someone else decides when it reaches staging, and a careful person runs the right scripts in the right order against production. It works, but only because people remember to do it right — in their heads, or in a spreadsheet.
EVOSCA takes that arrangement and puts it in the database. Changes are captured from the schema they were made in. Conflicts between developers are not discovered after the fact — they are detected early, before anyone's work is silently overwritten. Promotion to staging and production runs through approval gates that are rows in a table, not a convention. Versions are editions, so what is live is a property of the database rather than of a deployment log.
The name is what it does: Editioned Version and Schema Control Automation.
How it is built
Database-first
The catalogue, promotion state and conflict log are schema objects in the database being managed. Losing the tooling does not lose the history.
Tested against real engines
Changes are exercised against live Oracle and PostgreSQL instances, not mocks, through a simulation harness that runs multi-developer scenarios end to end before a release is cut.
Self-hostable by default
Built so that a team can run the whole thing inside their own boundary and read the code before it touches production. The managed service is the convenience, not the requirement.
Where it is up to
EVOSCA is not released yet. At first release the self-hosted build is free for up to 6 managed environments and 2 managed schemas: PostgreSQL as a public download, Oracle as early access on request. Hosted PostgreSQL opens at the same release, and the free tier is published while the paid tiers are still being worked out.
If you are running Oracle or PostgreSQL at a scale where any of this is familiar, the most useful thing you can do is try it and say where it falls short.
Get in touch
In the meantime, leave an address and we will reply.