Prep Library

AI / ML

From classical ML foundations to LLM systems — the concepts ML, MLE, and applied-AI interviews probe, each with intuition and tradeoffs.

Foundations

Overview

Foundations cover the core machine-learning concepts every practitioner is expected to reason about: the bias-variance tradeoff, regularization, choosing evaluation metrics, and engineering features. These fundamentals decide whether a model actually generalizes rather than just fitting the training data.

Bias-variance tradeoff

Under- vs over-fitting.

Core

Regularization

L1/L2, dropout, early stopping.

Core

Evaluation metrics

Precision, recall, F1, AUC.

Core

Feature engineering

Encoding, scaling, leakage.

Core

Deep learning

Overview

Deep learning covers how neural networks learn and the architectures behind modern AI — gradients and backpropagation, transformers and attention, and how to adapt pretrained models through fine-tuning, prompting, or retrieval. It's the basis for reasoning about today's large models.

Backpropagation & gradients

How nets learn.

Advanced

Transformers & attention

Self-attention at scale.

Advanced

Fine-tuning vs prompting vs RAG

Choosing an adaptation strategy.

Advanced

ML systems

Overview

ML systems is about running models reliably in production — MLOps practices for training, serving, versioning, and monitoring, plus rigorous evaluation of open-ended LLM outputs. This is what separates a notebook experiment from a dependable service.

MLOps & serving

Train, deploy, monitor.

Advanced

LLM evaluation

Beyond accuracy.

Advanced

Where this content comes from

For full transparency, this library's content is curated and verified from these sources:

Peer-reviewed ML papers & model cardsProduction ML engineering referencesOppZen-authored ML eval rubrics