Blog
SQL, data, and practical AI — plain English. Filter by topic or search.
All articles
Portfolio projects that do not look fake
Portfolio projects fail when they look like homework clones. Build one believable stack: real question, messy data, decisions, checks, and a short memo a hiring manager can trust.
Building a personal learning plan: SQL to Python to AI
A personal learning plan from SQL to Python to AI beats random tutorials. Build a sequenced path, a weekly hour budget, and proof of skill so you stop…
Airflow hello: retries, sensors, and when not to use it
Retries buy time for flukes; sensors encode waiting with deadlines. Learn when to retry, how to sensor a late file without burning workers, and when Airflow is the…
Churn and retention definitions that stick
Churn and retention sound like opposites until logo, revenue, and net metrics tell three stories. Lock definitions for who counts, what “active” means, and which window you report.
Access, least privilege, and just give me prod
“Just give me prod” is a speed request wearing a risk costume. Part 3 of Data stewardship at work maps least privilege from samples to write access, with…
