Building a Flight Data Pipeline Without Trusting AI

A few months ago I was cleaning up my GitHub and found this repo -- https://github.com/dat-a-ish/data-gov . It was built at that weird inflection point -- AI tools existed, Copilot was already in my editor, but I still rewrote every suggestion by hand because I didn't fully trust what came out. So the code in here is genuinely mine, typed line by line, and that's exactly what makes it useful as a learning resource. The irony? The project pulls real flight data from the Israeli Government API -- Ben Gurion Airport departures and arrivals. With the current situation in the Middle East making flights in and out of Tel Aviv a daily news story, a pipeline that tracks exactly that data feels more relevant than ever. But jokes aside -- this repo is not about geopolitics. It's a study in how a production-grade batch data pipeline actually fits together, from raw API ingestion all the way to a live dashboard. My honest recommendation: clone it, read it, and then build your own version without A