About

Before I wrote code for a living, I was a Human Intelligence Collector in the U.S. Army. The job was mostly talking to people — some of whom wanted to talk to me, and a lot of whom didn't. You learn a lot about what people actually mean versus what they say.

I moved into Data Engineering and then AI work, and to my surprise the Army job was better preparation for it than anything else I'd done. The hardest part of building anything at work isn't the code. It's figuring out that Jerry in accounting doesn't actually need the schema he asked for — he needs the numbers to line up a certain way when he opens the report Monday morning, and nobody ever told him what a schema was.

The thing I care about is getting people what they need to do their job, and making it easier and faster along the way. For years that meant being the bridge as a Data Engineer — connecting people with questions to the data that has their answers. Now, as an AI Engineer, it means building tools so they can find those answers themselves. Someone who finds their own answer has a deeper connection to the decision they're making. They own their part of the puzzle.

Day-to-day at work is mostly Python and getting people the data they need. At home I'm building ML inference pipelines and experimenting with the latest and fastest way to do things. I'm also working with management to bring more AI tools into our projects — I think we can do much more with AI than we currently are, so long as we use it the right way. It's not great for everything. Knowing when to reach for it and when not to is the distinction that matters.

Outside work I play a lot of poker. It combines math with reading people — basically the same job I do during the day, just with fewer spreadsheets.