100+ tabs open.
Zero plan.
It’s time to SAADHO.

Deep-ML · 30 min

Implement Self-Attention Mechanism

“Implement attention from scratch” is one of the most common ML coding prompts. This is the exact drill, with tests.

Skip: Don't optimise; correct shapes and a clean softmax are what's graded.

Anthropic · 20 min

Introducing Contextual Retrieval

A compact, production-tested walkthrough of hybrid search + reranking, with measured gains: great material for a RAG design answer.

Skip: The appendix tables.

hamel.dev · 25 min

Your AI Product Needs Evals

The reference text for how practitioners think about evals; interviewers often borrow its vocabulary.

Skip: “Eval Systems Unlock Superpowers For Free” on a first pass.

Hugging Face blog · 30 min

How to generate text: using different decoding methods for language generation with Transformers

“What do temperature and top-p actually do?” is a standard warm-up question, and this post builds each method on the previous one with runnable examples.

Skip: The “Appendix” links. Don't try to match the printed GPT-2 outputs word for word, and expect a couple of typos in the code comments.

OpenAI API docs · 40 min

Function calling

Almost every applied LLM loop asks you to explain or code the request → tool call → execute → result round trip, and this guide covers each knob interviewers probe.

Skip: “Custom tools” and all its grammar subsections (Lark/Regex CFG), “Defining namespaces”, “Tool search”, and “Streaming” unless you're building a chat UI.

Tech Interview Handbook · 40 min

Behavioral interviews for Software Engineers: How to prepare

Behavioral rounds weigh on hire and level decisions, and the usual failure is a rambling story with no clear personal action or result.

Skip: “Try out mock behavioral interviews” and “Expand your knowledge with newsletters, books, and courses”, which are mostly paid-product pointers.

NumPy documentation · 45 min

NumPy: the absolute basics for beginners

Interviewers judge NumPy fluency by whether you reach for shape, axis and broadcasting instead of a nested loop, and this page is where that vocabulary is defined.

Skip: “How to save and load NumPy objects”, “Importing and exporting a CSV” and “Plotting arrays with Matplotlib”; they never come up in a coding round.

jalammar.github.io · 40 min

The Illustrated Word2vec

“How are embeddings trained?” is the standard warm-up before any RAG or retrieval question, and word2vec with negative sampling is the canonical answer interviewers expect you to sketch.

Skip: “Personality Embeddings: What are you like?” if you already have the vector intuition, and “References & Further Readings”.

Google Machine Learning Crash Course · 50 min

Production ML systems

Batch versus online inference is usually the first serving decision an interviewer asks you to defend, and this module sets out its trade-offs in cost, latency, coverage of rare inputs and checking predictions before they ship.

Skip: “Monitoring pipelines” and “Questions to ask” for now; come back to “Monitoring pipelines” when you study ml-monitoring-drift.