Day 41 of 60 · Process, code review & AI-assisted

Code review (peer or AI)

Capers Jones puts review at 50–60% defect-removal efficiency, among the highest of any single technique. The cost is calendar time; the return is design errors caught before they ship.

ProblemDesign-level defects, missed requirements, and ambiguous specs that no static tool can catch.

How it works

Mandatory review of every change by another engineer or AI agent. Catches what the author can't see. Quality of review >> volume of review.

What it catches

Design errors, mis-specifications, missed edge cases. Capers Jones puts review at 50–60% defect-removal efficiency, among the highest.

Tools

Azure Repos PR · Hybrid GitHub PR · Hybrid GitLab MR · Hybrid Bitbucket PR · Hybrid Reviewdog · OSS CodeRabbit · SaaS Cursor / Claude Code · SaaS Graphite · SaaS

Verdict by project size

Small

Rec

Medium

Must

Large

Must

Extra-large

Must

Cost

Project size	Setup	Maint / mo	Tool / mo	CI / run
Small <10k LOC	0h	5h	$0	,
Medium 10–100k LOC	0h	25h	$0	,
Large 100k–1M LOC	0h	120h	$200	,
Extra-large >1M LOC	0h	500h	$5k	,

Setup = engineer-days to first useful run · Maint = engineer-hours / month at steady state · Tool = out-of-pocket $ / month · CI = minutes added (or saved) per pipeline run

Lifecycle & ownership

When in lifecycle

Plan Design Code

Per pull request · Runs in CI on every PR; gates merge.

Who owns it

Tech Lead / EM

DORA, post-mortems, ADRs

Collaborates with: Developer, Security / AppSec

Reference implementations

reviewdog
Automated PR review comments from linters and analyzers.
Danger JS
Automated pull-request checks and review guidance as code.
CodeRabbit docs
AI-assisted code review workflow with configurable review behavior.

Quick check

Capers Jones puts code review at roughly what defect-removal efficiency?

One question. Pick the best answer. Your streak is saved locally on this device.

Save the lesson

Download SVG ↓

Screenshot for a 1:1, drop it in Slack, or download the SVG.

All 60 days →