STAR CHARTS

Resources

An annotated reading list for people crossing the same bridge I am: infrastructure and DevOps folks pivoting toward data science and MLOps. Not a link dump - a shortlist, with a note on why each one matters when you come from the machine room.

⛩ The bridge - DevOps → MLOps

  • Hidden Technical Debt in Machine Learning Systems Sculley et al. · NeurIPS 2015 · paper

    The paper that explains why ML systems rot: the model is a tiny box in the diagram - the other 95% is infrastructure you already know how to build.

  • Rules of Machine Learning Martin Zinkevich · Google · guide

    43 rules, and the first one is "don't be afraid to launch a product without machine learning". Ops pragmatism, applied to models.

  • Designing Machine Learning Systems Chip Huyen · O'Reilly · book

    The system view: data, deployment, monitoring, drift. Reads like an SRE book that happens to be about ML - the fastest mental mapping for infra people.

  • MLOps Zoomcamp DataTalksClub · free course

    Hands-on and free: experiment tracking, orchestration, deployment, monitoring. Your CI/CD reflexes translate almost one-to-one.

  • ml-ops.org INNOQ · reference site

    Patterns and principles - the closest thing MLOps has to a twelve-factor manifesto.

🔭 Foundations - stats & ML

  • An Introduction to Statistical Learning James, Witten, Hastie, Tibshirani · free book

    The statistical backbone, free in PDF, with R and Python editions - pairs naturally with a TÉLUQ data science curriculum.

  • Hands-On Machine Learning Aurélien Géron · O'Reilly · book + notebooks

    From scikit-learn to deep learning with runnable code. The notebooks double as a reference you'll keep grepping.

  • StatQuest Josh Starmer · videos

    Every ML concept, dismantled one small step at a time. When a Kaggle metric confuses you, this is where you go first.

  • Neural Networks: Zero to Hero Andrej Karpathy · videos + code

    Backpropagation to GPT, built from scratch in code. The closest ML gets to "read the source, understand the system".

🛰 The platform - data & pipelines

  • MLflow documentation Databricks / Linux Foundation · docs

    Experiment tracking and the model registry - the artifact repository of the ML world, and the backbone of MLOps on Databricks.

  • Fundamentals of Data Engineering Joe Reis & Matt Housley · O'Reilly · book

    The data lifecycle end-to-end, refreshingly tool-agnostic. Explains where your pipelines sit in the bigger picture.

  • Kaggle Learn Kaggle · free micro-courses

    Short, practical, immediately applicable in competitions - the fastest feedback loop between theory and a leaderboard.

// curated by a human, revised as the journey progresses - suggestions welcome via GitHub