Cloud platforms across AWS, Azure and GCP

Context

At Webloo, my responsibilities spanned the architecture and implementation of cloud platforms across AWS, Azure and GCP. The work brought application requirements, delivery processes and operational responsibilities into the same set of engineering decisions.

Across those projects, the questions went beyond where an application should run. They included how environments were provisioned, how services communicated, how access was controlled, and how a team would understand a failure after deployment.

My responsibility

As Software Engineering Lead and DevOps Architect, I led the design and implementation of cloud solutions and worked with cross-functional teams on their delivery. My scope included infrastructure automation, deployment architecture, security controls, observability and the connection between platform choices and business requirements.

That work built on my earlier DevOps Team Lead responsibilities at Webloo, which included multi-cloud architecture, infrastructure provisioning and full-stack project delivery.

Engineering approach

The cloud services in this work included EC2, S3, ECS, EKS and CloudFront on AWS; BigQuery and Cloud Functions on GCP; and App Services and Azure SQL on Azure. These were tools used across the work, with their responsibilities determined by the application and operating context.

I used Terraform and Ansible to make infrastructure and configuration changes repeatable. IAM policies, encryption and secrets management with HashiCorp Vault formed part of the security work. For operational visibility, I worked with OpenTelemetry, Prometheus, Grafana and Azure Monitor.

I also treated documentation and team coordination as part of the platform. Infrastructure needs to be understandable to the people developing, deploying and operating the applications it supports.

Complexity and trade-offs

Working across cloud providers means dealing with different identity models, networking behavior, deployment interfaces and managed-service constraints. A common engineering approach needs to preserve those differences rather than hide them behind a diagram.

Cost, fault tolerance and integration were part of the design work. Those concerns can pull in different directions: more redundancy adds cost and coordination, while additional managed services introduce operational dependencies. I worked through those decisions alongside the requirements of each project.

What this work brought together

My professional experience here combines cloud architecture with implementation: infrastructure as code, application hosting, security controls, deployment automation and operational visibility. It is the point in my work where software design and the environment running that software have to be considered together.

Sources & further reading

• Professional profile