I build backend systems, cloud infrastructure, and developer tooling, mostly in Go. I’m Douglas Adolph, and I work from Austin, Texas. I care about software that is dependable, maintainable, and straightforward to operate.
What I do
Eight-plus years across software, platform, and systems engineering. My official title is senior software engineer, but as the only Go developer on a Kubernetes platform team at enterprise scale, the job runs wider than the title: platform engineering, DevOps, and systems work, plus the Go that only I write. It covers deployment, identity, secrets, observability, and the tooling that keeps large systems boring to run.
The scale behind that: a fleet of 2,300+ production clusters, plus hundreds more for testing. The deployment service I designed and own took releases to that fleet from weeks or months down to seconds. I own half a dozen load-bearing services, and more than a dozen tools I have written are in use across the company.
On my own time
I build software from scratch. Command-line tools, backend services, web applications, libraries, specifications, and open source. I like taking an idea from an empty repository to something people can actually use.
Before software
I came to engineering late. The decade before it was service, retail, and sales work. Kitchens and grocery aisles, waiting tables, a shipping hub, commission sales, a pawn shop, and years on a call center phone queue that ended with me managing the floor.
Every one of them taught me something I still use.
I went back to school for computer science at Texas State, commuting to San Marcos from Austin with two kids at home and a third born partway through the degree. I went into it for the financial security and did not expect to end up loving it.
How I work
Standard-library-first Go and a serious testing discipline. I would rather automate a recurring problem than be good at handling it. The best compliment my software gets is that nobody thinks about it.
I ship single binaries wherever I can, with the assets embedded and no build step to rot. When something needs a web interface I will usually vendor one library and hand-write the rest, which sounds like more work and is reliably less. Software with no toolchain still runs in five years.
I use AI extensively in my development workflow, but I treat it like any other engineering tool, useful only when paired with judgment, testing, and review. I wrote up how that actually works, including the rules I do not break.
Tools of the trade
Go, Kubernetes, Terraform, HashiCorp Vault, Google Cloud and GKE, Prometheus, Docker, and whatever else the problem needs, including the front end.
Away from the keyboard
Husband and father of three, plus a Labrador. Writer, musician, tinkerer, outdoorsman, and a reader of history, philosophy, and science.
Education
B.S. in Computer Science and Engineering, Texas State University, with a minor in Applied Mathematics.
Elsewhere
Reach me at [email protected]. See what I take on at consulting, or read how I use AI.