Depending on how we setup the vector database and it's data ingestion, we may also need some way to update the datastore with each new deploy (or deploys to a specific, data related branch?)
Merges to develop should deploy to a dev environment/stack in AWS
Merges to main should deploy to a staging environment/stack in AWS
Creating a release tag on main should deploy to a production environment (ideally the release tag is included in the stack metadata or resource names)
Pushes to feature branches (feature/*) should trigger a deployment to a similarly named environment/stack in AWS, run unit and integration tests against that stack. Merging the feature pranch into develop should destroy the previously deployed feature stack in AWS.
developshould deploy to adevenvironment/stack in AWSmainshould deploy to astagingenvironment/stack in AWSmainshould deploy to a production environment (ideally the release tag is included in the stack metadata or resource names)feature/*) should trigger a deployment to a similarly named environment/stack in AWS, run unit and integration tests against that stack. Merging the feature pranch into develop should destroy the previously deployed feature stack in AWS.