First episode of what I’m calling Mert Talks Data — short, practical episodes about data engineering and ML in the real world.
This pilot covers how I built ozamert.com: the Astro + MDX stack, why I chose it over a Notion-hosted blog, and the AI agent workflow I set up to draft posts with local LLMs (Qwen2.5-Coder via Ollama).
What’s coming
Future episodes will follow the same format as the blog: show a real problem, walk through the fix, be honest about dead ends. No padded intros, no sponsor reads until there’s a real audience.
If you’re reading this before there’s a proper podcast feed — hi, you found it early.