# Overview JadeGit provides the ability to [build deployments](https://jadelab.gitbook.io/jadegit/0.15.0/console/build), which may be published/run in different ways. * Using an automated pipeline on a platform like Azure DevOps to push changes to target databases. This approach would be suitable in scenarios where all databases can be accessed/managed by the platform (i.e. using [environment resource agents](https://learn.microsoft.com/en-us/azure/devops/pipelines/process/environments-virtual-machines) to manage a database). * Publishing updates to a feed that can be queried/pulled from by remote databases. This approach would be suitable in scenarios where there are many databases across client sites which cannot be accessed/managed directly, although publishing to the feed could still be performed by an automated pipeline. The fundamental steps for building & deploying updates remain the same irrespective of the direction or using an automated process. 1. Retrieve the current commit ID for the database (or feed). 2. Build deployment, supplying current/previous commit ID and latest commit ID. 3. Run deployment against database. 4. Update the current commit ID, ready for retrieval next time. To publish updates to a feed instead of pushing updates out directly, a staging database could be used to confirm deployments work before they're added to the feed (which could be organized/queried by commit ID). As it'd represent updates previously published, the current commit ID could be retrieved from the staging database during the first step above. The current commit ID for a database may be stored as preferred (i.e. could just be a simple text file), or you may use the JadeGit [registry](https://jadelab.gitbook.io/jadegit/0.15.0/deployment/registry) (recommended).