// Articles
Guides, tutorials, and comparisons to help you build better backend applications.
Design the backend that sends 24-hour and 2-hour appointment reminders exactly when they are due, never twice, in the clinic's time zone and without medical details in the message, built by hand on AWS or in TypeScript with Encore.
Design the backend behind fintech customer onboarding: collect identity documents through signed uploads, start one verification check per application, accept signed provider webhooks and keep an auditable status trail, built by hand on AWS or in Go with Encore.
Terraform for a private Cloud Storage bucket with versioning, soft delete, lifecycle rules and CORS, IAM that lets a service create but not delete objects, and the permission signed URLs need on Cloud Run, plus how Encore provisions buckets from code.
A complete Terraform setup for an API on Amazon ECS Fargate behind an Application Load Balancer, with security groups, logs, deployment rollback and CPU autoscaling, the settings that cause failed deployments, and how Encore deploys to Fargate from code.
Terraform for Secrets Manager secrets whose values never reach state, a customer managed KMS key, read access for an ECS service, rotation options, and the deletion behavior that breaks re-applies, plus how Encore handles secrets from code.
Terraform for a Pub/Sub topic and pull subscription with retry and dead-letter policies, the IAM bindings that dead lettering quietly depends on, and how Encore provisions the same topics from Go or TypeScript.
Terraform for a Cloud SQL for PostgreSQL instance with private IP, backups and point-in-time recovery, a database user whose password never lands in state, and the settings that cause trouble later, plus how Encore provisions Cloud SQL from code.
Which IAM role on an Amazon ECS task definition does what, which permissions belong on each, the errors you get when they are swapped, and Terraform that sets both up for a Fargate service.
How Google's two serverless platforms differ on runtimes, scaling, timeouts, background work and bundled services, when App Engine still makes sense, and what a move to Cloud Run involves.
Move from Heroku Key-Value Store (formerly Heroku Redis) to ElastiCache in your own AWS account: sort keys into what to drop, drain and copy, then carry live sessions across with their remaining TTLs.
Move a Fly Managed Postgres or unmanaged Fly Postgres database into Amazon RDS in your own AWS account, using flyctl proxies, pg_dump and pg_restore, then point a Go backend at the new database.
Move a Railway Postgres database into Amazon RDS in your own AWS account through Railway's public TCP proxy with pg_dump and pg_restore, with a rehearsal, a short write freeze and checks before cutover.