Section 01
ML System Architecture Fundamentals
0 / 4
Continue Here
From Notebook to Production
Easy
Available
Batch vs. Online, Not Both at Once
Easy
Available
A Safe Path to Raw Data
Medium
Available
A Small, Complete ML Lifecycle
Medium
Section 02
Data Collection & Labeling
0 / 4
Available
Sampling Bias in Event Collection
Easy
Available
Joining Late-Arriving Labels
Medium
Available
Human Labeling Throughput
Medium
Available
Filtering Sensitive Data Before Training
Medium
Section 03
Data Pipelines & Preprocessing
0 / 4
Available
Outgrowing the Nightly Window
Medium
Available
A Realtime Feature Arrives Too Late
Medium
Available
Duplicate Events Corrupt Training Examples
Medium
Available
Backfills Overwhelm the Pipeline
Medium
Section 04
Feature Engineering & Feature Stores
0 / 4
Available
Training-Serving Skew
Medium
Available
Online Feature Lookups Under Peak Load
Medium
Available
Freshness SLAs Differ by Feature
Medium
Available
Six Teams, One Feature
Hard