Hakopod
Builder & lead contributor · Current
An open-source application platform built around resumable deployments, isolated services and recoverable database workflows.
Go · PostgreSQL · Kubernetes · Native database operators
Deployments that can recover
I built a Go control plane with durable PostgreSQL queues, revision checks and idempotent retries. Deployment work can resume after a client disconnects or a service restarts, rather than depending on a single uninterrupted request. Kubernetes service isolation and private networking keep applications separated, while rolling deployments provide a controlled release path.
Recovery is part of the workflow
Failed-release recovery and an auditable rollback history are built into the platform. I also added database provisioning and recovery workflows using native operators, scoped credentials and verified TLS connections. Restores use separate targets, keeping recovery work distinct from the original database.
An ongoing open-source platform
Hakopod is current work, with my role spanning the control plane, deployment lifecycle and database workflows. The engineering focus is durable operations: work that can resume, releases that can be rolled back, and recovery paths that can be inspected.
