the archive
Articles
150 deep dives, ~32 hours of hands-on builds. Filter by topic, or press ⌘K to search everything.
150 articles
30 Days of DevOps
30 articles
Git from Zero — Your First Commit, Branches, and the Two Workflows Real Teams Use
Git tutorial for beginners: install Git, make your first commit, and learn branching hands-on — then compare GitFlow vs trunk-based development, the two workflows real DevOps teams use.
Dockerize Any Application the Right Way — Multi-Stage Builds & Best Practices
Docker multi-stage build tutorial: shrink a 1.2 GB image to 47 MB with a distroless, non-root Dockerfile, and verify the security gains with Docker Scout. Dockerfile best practices.
Docker Compose for a Full Local Dev Stack — Node.js, PostgreSQL, Redis, and Nginx
Docker Compose tutorial: run a full local dev stack — Node.js REST API, PostgreSQL, Redis, and Nginx — wired together with health checks, named volumes, and startup ordering.
GitHub Actions CI/CD — Automated Build, Test, Scan, and Push on Every Commit
GitHub Actions CI/CD tutorial: build a 3-job pipeline that tests your Docker image, blocks bad merges, scans for CVEs, and pushes to GHCR on every commit — in under 2 minutes.
Kubernetes Fundamentals — Local Cluster with kind, Pods, Services, and Zero-Downtime Rollouts
Kubernetes tutorial for beginners: run a local cluster with kind, deploy a containerised app, add liveness and readiness probes, and ship a zero-downtime rolling update — no cloud needed.
Helm — Package Manager for Kubernetes. Charts, Templates, and Multi-Environment Releases
Helm tutorial for Kubernetes: turn raw YAML into a reusable Helm chart, configure dev vs prod with values.yaml, then install, upgrade, and roll back releases with single commands.
Python for AI Engineering
30 articles
Set Up Python on Any OS and Write Your First Program
Day 1 of Python for AI Engineering. Install Python on macOS, Windows or Linux, set up VS Code and your first virtual environment, then write and run a real Python program — an interactive intro-card generator. No prior coding experience needed; everything runs on your own laptop.
Variables, Numbers, Strings & Booleans
Day 2 of Python for AI Engineering. Master the raw materials of every program: variables, the four core data types (int, float, str, bool), arithmetic, string formatting, comparisons, and the type conversions that make user input usable. You'll build an interactive tip & bill-splitting calculator.
Lists, Tuples, Sets & Dictionaries
Day 3 of Python for AI Engineering. Meet Python's four built-in collections — lists, tuples, sets and dictionaries — learn exactly when to reach for each, and the operations you'll use daily. You'll build a contact book that uses all four together.
Loops & Conditionals
Day 4 of Python for AI Engineering. Make your programs decide and repeat: if/elif/else, the and/or/not logic, for loops over collections, while loops, and break/continue. You'll turn the contact book into a real menu-driven app you can use again and again.
Functions & Return Values
Day 5 of Python for AI Engineering. Learn to package code into reusable functions: def, parameters and arguments, return values (and how they differ from print), default and keyword arguments, and multiple return values. You'll refactor the contact book so every menu action becomes a clean function.
Scope, Imports & Modules
Day 6 of Python for AI Engineering. The last piece of the foundations: where variables live (scope), how to split code across files (modules), how to import your own code and Python's huge standard library, and the if __name__ == '__main__' idiom. You'll split the contact book into a reusable module and a main program.
100 Days of MLOps
90 articles
What is MLOps, and Setting Up Your Machine on Any OS
Day 1 of 100 Days of MLOps. Understand what MLOps actually is in plain English, then install the four tools every ML engineer needs — Python, Git, VS Code and Docker — on macOS, Windows or Linux. You finish by running a real environment-check program that proves your machine is ready. No CS background needed; everything runs locally, no cloud, no bills.
The MLOps Lifecycle & Mental Model
Day 2 of 100 Days of MLOps. Before we touch a single model, get the map. Learn the loop every ML system lives in — data, train, package, deploy, monitor, retrain — in plain English, see exactly where each tool fits, and write a one-page project charter for the house-price predictor we'll grow across the series. Runs 100% locally on any OS.
Python Environments for ML: venv, pip & requirements.txt
Day 3 of 100 Days of MLOps. Stop turning your laptop into a dependency minefield. Learn virtual environments (venv) — isolated, throwaway workspaces — install your first real ML libraries (NumPy, pandas, scikit-learn), and pin them in requirements.txt so anyone can rebuild your exact setup from scratch. Your first reproducibility muscle. Works on macOS, Windows and Linux.
Git Basics for ML Projects (and a Proper .gitignore)
Day 4 of 100 Days of MLOps. Give your project a memory. Learn the everyday Git commands — init, status, add, commit, log — by putting the house-prices project under version control, and write an ML-aware .gitignore so you never commit the things that don't belong in Git: your bulky .venv, giant datasets, trained models and secrets. Works on macOS, Windows and Linux.
Notebooks vs Scripts: Explore Fast, Ship Reliably
Day 5 of 100 Days of MLOps. Jupyter notebooks are where ML ideas are born; clean .py scripts are what actually ships. Learn both, understand the notebook 'hidden state' trap, and master the real MLOps workflow: prototype in a notebook, convert it with nbconvert, then clean it into a script a pipeline can trust. Works on macOS, Windows and Linux.
Your First ML Model, End to End
Day 6 of 100 Days of MLOps. The moment it becomes machine learning. Train a real model with scikit-learn on the house-prices dataset — load the data, split it honestly, fit a model, and measure how well it predicts with MAE and R². You'll understand every line and every number. Runs 100% locally on macOS, Windows and Linux.