The language keeps changing; the kind of problem I like hasn't. Across every job I've had, I've worked the same layers — the backend, the data, the parts where being wrong has consequences — and I've changed tools each time to keep doing it.

The path here

I started in C, writing Unix middleware. When the web arrived I moved to Java and built web applications; somewhere in there I spent a stretch troubleshooting DB2 on the mainframe, which taught me more about data under pressure than any course could. Later came full-stack work in C#, .NET, and SQL Server, with a Vue front end.

Somewhere in the middle I went back for a Computer Science degree and a math minor, and worked as a TA. The course I most enjoyed teaching was intro programming in Python, which is also where I taught myself the language. Each of those shifts was the same move: a new stack, the same instinct for what I wanted to build with it.

Where I work best

My centre of gravity is the backend and the data tier — schema design, the logic that has to be correct, the parts a user never sees but feels when they're wrong. That's the work I reach for and the work I'm best at.

I'm comfortable at the front end too, and more so lately: GenAI has turned the interface layer into something I can move through quickly to get to the part I actually care about. I'd rather use good tools to build a clean UI fast than spend a career polishing one.

Now

Python is what I work in by choice. I'm most interested in where it meets data that carries weight — financial applications, and the newer work around AI agents. Those are less different than they look: both come down to getting the data right, keeping the architecture honest, and knowing when to distrust the output.

The clearest example is the most recent thing I built, a Canadian tax estimator — as much an experiment in whether AI-assisted development can produce genuinely maintainable code as it is a tax tool. Building it, I found that the writing around the code — the README, the docs that explain why a thing works the way it does — is a part of the craft I take seriously, and increasingly one I enjoy in its own right.

How I work

The last few years have been quieter on paid work and busier on family and on study — a stretch of deliberate retooling into Python, data, and AI. Some of what came out of it: ten DataCamp tracks across Python, data analysis, and data engineering, and four Coursera specializations in generative AI and prompt engineering.

I work remotely or hybrid, and that's a firm requirement tied to family care commitments I'm not looking to change. It's non-negotiable, so I'd rather say it early than waste anyone's time.