Executive Summary

Scrum, as practiced in most organizations, is broken. The word has lost its meaning. Every company and every team interprets it differently, and the gap between what the Scrum Guide describes and what teams actually do has become so wide that saying "we do Scrum" communicates almost nothing about how a team works.

At the same time, AI is joining software teams as a participant, not just a tool. Whether through coding assistants, autonomous agents, or embedded AI features, a growing share of implementation work is now shared between humans and AI systems. The same team can now deliver dramatically more, and that amplified capacity changes what coordination looks like. Scrum, designed to coordinate human-only teams producing at human-only speed, has no model for this new reality.

Dandori is a team coordination framework built for this convergence. Named after the Japanese manufacturing concept meaning "preparation" or "setup," Dandori reflects the core insight of the Toyota Production System: the quality of output is determined by the quality of preparation, not by what happens on the production line.

In Dandori, the specification is the unit of work. Human judgment is the scarce resource. AI is the execution engine, though humans remain in the loop across the full spectrum from AI-assisted to fully human-implemented work. The framework provides a complete operating model: a five-stage specification lifecycle, three defined roles, a product alignment model, a redefined engineering management function, and a set of ceremonies that transform rather than eliminate Scrum's core practices. Each element exists to optimize the flow of human thinking through a system where machines handle an increasing share of production.

"Scrum optimized for production. Kanban optimized for flow. Dandori optimizes for preparation, because when AI handles production and flow is continuous, the quality of what you prepare determines everything."