Luke Lesurf
Senior platform engineer in Hertfordshire. I build and run the infrastructure other engineers ship on: AWS, Kubernetes, Terraform, CI/CD, and the observability to know when it is lying to you. Twenty years of software before that, mostly Java. Security background from a stint in penetration testing.
Outside work I run a flight data warehouse on a home server that ingests about 37 million ADS-B position reports a day and keeps 9.2 billion of them queryable. That is the hobby. It is also the best description of how I think about systems, so it gets a post.
How I work
Infrastructure is a product
The people using the platform are customers. If they need a ticket to deploy, the platform has failed.
Honest about gaps
I say what I have not done. It is cheaper than finding out in production.
Measure, then optimise
Look at what the box is actually doing. Remove the thing it does most often for no reason.
Agents do the typing
Spec and tests first, coding agents do the implementation, I do the reviewing. It is how I build everything now.
Now
Platform enablement at a large payments company: migrating legacy applications onto an internal PaaS, rebuilding their CI/CD, writing the Terraform the platform does not cover yet. At home, adding worldwide METARs and airband transcription to the flight data system, and finally writing some of it down here.
Writing
- A flight data warehouse on a home server
9.2 billion ADS-B position reports, 46 million flight legs, one box with 16 threads. How it is built and what it taught me about keeping and throwing away data.