Profile
I am a Cloud DevOps engineer working on Azure and AWS. I joined Kortext, a higher education platform with three million users, as its only dedicated and specialist infrastructure engineer in August 2024. I brought the estate under Terraform and Azure DevOps pipelines, built out the monitoring and on-call function, cut cloud spend by a third, and helped build the platform team of five that now runs it. I am currently working alongside Kortext's AI development team to build and support Kortext IQ, the company's newly launched AI assistant for higher education.
How I think
On cost, I do things in a fixed order: visibility, right-sizing, scaling, then reservations. Tag everything and put the spend in front of the people who own it. Size from utilisation data. Add autoscaling so the estate follows demand. Only then commit to reserved capacity. Most people buy reservations first, which locks in whatever was already oversized.
On incidents, the first fifteen minutes are mostly dead time while the on-call engineer finds a laptop and gets their bearings. I built an agent to use that window. By the time they are in, there is a written investigation with candidate causes waiting for them instead of an empty dashboard. It deliberately runs away from the estate it investigates, so it still works when that estate is what has broken.
On ingress, my instinct is to reduce the number of ways in. Kubernetes will happily give every service its own public load balancer, and DNS-layer routing only hands out an address, so the origin stays reachable by anyone who resolves the name. I would rather have one Layer 7 entry point that traffic has to pass through, on Application Gateway with AGIC. That gives you one place to terminate TLS, apply a WAF and rate limits, and see what is actually arriving, with far fewer public addresses to defend. Security and control are worth the extra hop.
