arrow_backKnowledge Store

Category 02

Architecture

7-layer hierarchy model, 3-role model (A-B-C), concurrent task model, event-driven architecture — Structural design for AI-native organizations.

7

Hierarchy Layers

3

Role Model (A-B-C)

4

Core Principles

0

Zero Polling

Core Documents

Hierarchy spanning from purpose to implementation, role allocation, concurrency, and event-driven design — the blueprints defining the system's backbone.

02-01 stacks

7-Layer Hierarchy Model

7-layer decomposition from Higher Purpose to Subtask. AI-driven top-down decomposition and consistency checking.

1 Higher Purpose
2 Mission
3 Goal
4 Initiative
5 Epic
6 Story
7 Subtask
Asana GitHub Immutable
02-02 group_work

3-Role Model (A-B-C)

Autonomous decentralized organization design with 3 roles: Framer, Orchestrator, and Executor.

A = Framer

The visionary who defines problems and sets the vision

B = Orchestrator

AI agent that decomposes tasks and manages flow

C = Executor

Fast-executing AI agents + specialists. "Compound interest of AI speed"

A-B-C Event-Driven Compound Effect
02-03 call_split

Concurrent Task Model

Concurrency design through 2-axis transformation: skill-dependent to process-dependent, synchronous to asynchronous.

swap_horiz

2-Axis Transform: Skill-dependent to process-dependent, sync to async

merge_type

Principle: "Structured merges enable unstructured work"

pattern

3 Patterns: Fully independent, branch-merge, hierarchical concurrency

Async Branch-Merge Optimistic Lock
call_split
02-04 electrical_services

Event-Driven Architecture

Real-time system powered by Event Stream + Knowledge Graph + Materialized Views + Actor Groups.

stream

Core Components: Event catalog and TypeScript type-safe design

flash_on

4 Principles: Zero Polling, real-time propagation, concurrent evolution, AI speed compound interest

Tech Stack

Kafka Event Stream
Neo4j Knowledge Graph
PostgreSQL Materialized Views
Kafka Neo4j TypeScript