I’ve spent the last year building AI-powered automation pipelines for digital marketing — search term classification, automated reporting, creative performance audits, predictive LTV modeling — and I’ve learned more about engineering in that time than in the previous five years of my career.
This site is where I start sharing what I build.
What you’ll find here
Every project I publish comes with a CLAUDE.md file. That means you can clone the repo, point Claude Code at it, and start adapting it to your own setup. No need to understand every line of code on day one — that’s the point.
Topics I’ll be writing about:
- Search term classification with AI — using Gemini and Tavily to automatically classify and route Google Ads search terms
- Automated marketing reports — generating MBR/WBR reports from BigQuery data with AI-written commentary
- Creative performance auditing — scoring ad creatives across channels using structured naming conventions
- The Claude Code workflow — how a marketer (not an engineer) uses AI to build production pipelines
Why share this?
Because the best way to learn is to build, and the second best way is to explain what you built. And because I wish someone had shared this kind of thing when I was starting out.
More soon.