Multicloud without architecture religion
A practical view of choosing monoliths, services, serverless, local infrastructure, private cloud, public cloud, or hybrid models based on the product and operating context.
Blog
Notes on product operations, private cloud, hybrid systems, DevOps support, delivery pipelines, observability, cost control, security, and cloud architecture tradeoffs.
Articles
A practical view of choosing monoliths, services, serverless, local infrastructure, private cloud, public cloud, or hybrid models based on the product and operating context.
How to decide whether a product belongs on private cloud, using workload shape, data control, cost predictability, operating capacity, and migration risk.
Hybrid cloud can support real products when identity, network paths, deployment, observability, support handoff, and incident ownership are designed as one model.
Cheap infrastructure choices often move cost into incidents, manual work, poor observability, customer support pain, vendor lock-in, and slower product delivery.
What growing product teams need from DevOps support: safer releases, clearer environments, cost visibility, incident readiness, and operations without enterprise drag.
Kubernetes is powerful, but many products need simpler deployment, environment, rollback, and operations patterns before a full orchestration platform makes sense.
Reliable product delivery needs clear stages, controlled configuration, useful checks, rollback paths, release visibility, and fewer unnecessary platform layers.
Observability should show product behavior, failed jobs, customer-impacting errors, latency, and recovery signals, not dashboards nobody uses during incidents.
A practical framework for choosing public cloud, private cloud, or hybrid infrastructure from the product workload, support model, constraints, cost, and operating capacity.
Subscribe
The newsletter is built for people who make product, delivery, platform, support, and cloud decisions.