Current deployment takes 40+ seconds, we need to root cause and possibly improve the deployment and iteration speed. - [ ] Profile current deployments - [ ] Keep Kaniko running? - [ ] Use EC2 capacity provider instead of Fargate? - [ ] Local "Docker" builds for debug/development mode?