CD Smith logo

Architecture Decision Toolkit

An architecture pattern is a proven way to organize a system: where the code lives, how its parts talk to each other, and how it is deployed and scaled.

Choosing the right one up front is hard. There are well over a dozen common patterns, many of them overlap, and several are meant to be combined. Their costs often show up only after the system has grown, when changing course is expensive. The resources below help you narrow the options, compare trade-offs side by side, and explain your reasoning in an interview.

Choose a pattern

Two decision trees ask yes/no questions from opposite ends. Use whichever matches how you think about the problem, or try both and compare where you land.

Decision treeWeb page

Complexity-first

Starts with the most specialized needs, such as data pipelines, plug-ins and audit history, and works down to a simple layered monolith as the default.

Specialized → Data and communication → Overall shape Open the tree →
Decision treeWeb page

Simplicity-first

Starts with the simplest architecture and asks whether it meets the need. You only move to a costlier pattern when the answer is No.

One app → Add a pattern → Distributed → Specialized Open the tree →
Comparison matrixExcel

Pattern matrix

All 12 patterns side by side with team size, color-coded complexity, key features, pros and cons. Filter or sort by any column.

Download the spreadsheet →

Prepare for interviews

Read the guide first for when to use each pattern and what interviewers look for, then practice with the question-and-answer set.

Interview guidePDF

Architecture patterns: interview guide

A short overview of each pattern: when to use it, pros and cons, and the points worth raising in an interview. Ends with a quick comparison table.

Open the PDF →
Practice questionsPDF

Interview questions and model answers

About 45 likely questions with model answers for each pattern, plus scenario questions on choosing between patterns and migrating systems.

Open the PDF →