$ cat ./ml-prep/resources.md
ML Interview Prep
A curated, opinionated set of resources I reach for when preparing for ML interviews — organized by the buckets most loops actually test. Free and high-signal first.
# ML Breadth & Fundamentals
Core concepts you should be able to explain on a whiteboard: bias/variance, regularization, evaluation, classic algorithms.
by Chip Huyen
The canonical interview-prep reference. Question bank + how loops work.
by James, Witten, Hastie, Tibshirani
Free PDF. The best intuition-building book for classic ML.
- CS229 Machine Learningcourse
by Stanford
Lecture notes are gold for the math behind the algorithms.
- Google ML Crash Coursecourse
by Google
by Aurélien Géron
The most practical end-to-end intro — code-first coverage of classic ML through deep learning.
# Deep Learning
Neural nets, backprop, optimization, and modern architectures. Be ready to derive backprop and reason about training dynamics.
by Andrej Karpathy
Build backprop, an MLP, and a GPT from scratch. Essential.
Free, code-first, covers CNNs/RNNs/Transformers with runnable examples.
by Goodfellow, Bengio, Courville
by Jay Alammar
The clearest visual explanation of attention.
# Coding & Implementation
DS&A for the coding round, plus 'implement X from scratch' (k-means, logistic regression, attention).
- NeetCode 150site
Curated LeetCode path with clean explanations.
- LeetCodesite
by Erik Linder-Norén
Reference implementations you can study and reproduce.
# ML System Design
Designing end-to-end ML systems: data, features, modeling, serving, monitoring, and feedback loops.
by Chip Huyen
The reference for production ML and ML system design rounds.
by Alireza Dirafzoon
Framework + worked examples for the system design round.
by Ali Aminian & Alex Xu
Worked walkthroughs of the most common ML design problems (feed ranking, recs, search, ads).
by Google
Hard-won practical advice for ML in production.
# LLMs & Generative AI
Increasingly tested: transformers, fine-tuning, RAG, evaluation, and the tradeoffs of building with LLMs.
by Andrej Karpathy
- Hugging Face LLM Coursecourse
by Hugging Face
by Anthropic
Patterns and anti-patterns for agentic LLM systems.
- AI Engineeringbook
by Chip Huyen
Building applications on foundation models: evaluation, RAG, fine-tuning, and inference tradeoffs.
by Jay Alammar & Maarten Grootendorst
Visual, code-first tour of tokens, embeddings, transformers, and LLM applications.
# MLOps & Production
Taking models from notebook to production: deployment, containerization, experiment tracking, CI/CD, and monitoring.
- Made With MLcourse
by Goku Mohandas
End-to-end MLOps: design, develop, deploy, and iterate. Free and code-first.
- Full Stack Deep Learningcourse
Productionizing ML: infra, testing, deployment, and monitoring.
by Google Cloud
The canonical reference for MLOps maturity levels and CI/CD/CT.
# Search, Ranking & Recommendations
The bread-and-butter of applied ML at scale: retrieval, learning-to-rank, candidate generation, and recommenders.
by Google
Candidate generation, scoring, and re-ranking from first principles.
by Eugene Yan
Deeply practical writing on search, ranking, recs, and ML system design.
# Behavioral & Communication
Often underweighted by candidates, rarely by interviewers. Have crisp STAR stories ready.
Structure for answering 'tell me about a time...' questions.
by Will Larson
Useful for senior/staff-level scope and impact framing.