Skip to content

KARYON AI Architecture Book

Welcome to KARYON - The Architecture of a Cellular Graph Intelligence

Karyon is in active production-readiness work. The documentation describes the intended organism, but the current repository still contains a mix of implemented runtime, partial integrations, and explicitly mocked development paths.

  • The umbrella app layout, OTP supervision, and Rust NIF boundaries are real.
  • Memgraph, XTDB, NATS, and Firecracker integrations still need service-backed validation before being treated as production-ready.
  • Runtime contracts across core, rhizome, and nervous_system are being normalized under the root PLAN.md.
  • Use the root TASKS.md to see which phases are complete, in flight, or blocked.

Part I: The Biological Edge in Systems

Explore the core biological primitives required to build a reasoning, adapting organism. Read Chapter 1

Part II: Anatomy of the Organism

Delve into the Karyon microkernel and the technologies—Elixir, Rust, and KVM—that bring it to life. Read Chapter 3

Part III: The Rhizome (Memory & Learning)

Learn how the AI experiences, restructures, and optimizes memories inside a sprawling graph database. Read Chapter 5

Part IV: Perception and Action

Understand the boundaries between the organism’s internal reasoning and the chaotic external world. Read Chapter 7

Part V: Consciousness and Autonomy

The mathematical framework that elevates algorithms into curiosity-driven, self-optimizing entities. Read Chapter 9

Part VI: Maturation & Lifecycle Execution

The hands-on framework for training the 500k-cell colony, maintaining codebases, and isolating engrams. Read Chapter 11