Poesis
The Systemic Intelligence Engine — the governed AI context.
Reference implementation of GSM. What Kubernetes is to software infrastructures, the Systemic Intelligence Engine is to the world. Both are declarative reconciliation engines: you define the desired state as typed, versioned, API-managed objects, and the engine continuously evaluates the real world against that definition and acts to close the gap. From virtualized and managed computing resources to virtualized and managed world resources. This is the logical and natural evolution path as we enter the generative AI era. The key to this evolution path is the systemics model: GSM. This evolution enables generative AI to operate a capable and trustworthy synthetic autopoiesis.
Why it matters
The governed AI context — the layer context engineering is missing.
A governed AI context.
SIE leverages the definitions as a typed, machine-readable context — each one carrying its Archetype’s meaning — that applications, humans, and agents can reason and generate from directly. It is the governed account generative models rarely have — what a system is, what it must do, and why — the layer context engineering is still missing.
Emerges from Definition Manager
Harnessed operations.
Operations on that context are harnessed by the definitions themselves: Directives and Norms bound what AI and agentic actions may do, so every operation stays governed, checked, and trustworthy — intuition proposes, the model disposes.
Emerges from Definition Manager · Operator
One closed self-sustaining loop.
Define → realize → evaluate → refine: definitions are stored and lifecycle-enforced, executed against observed state, and refreshed from collaborative sourcing — the system continuously regenerates itself against its own definition, and every divergence surfaces immediately, not at audit time.
Emerges from Definition Manager · Operator · Definition Blackboard Manager
Reality feeds governance.
The Definition Blackboard Manager lets code sourcers, manual authors, and external frameworks contribute evidence-backed identifications — sealed, audited, and promoted into the Definition Manager’s governed lifecycle, so the model tracks the real system.
Emerges from Definition Blackboard Manager · Definition Manager
The mental model
One loop, from definition to reality — and back.
Think Kubernetes, generalized: you declare desired state as typed, versioned, API-managed definitions, and the engine continuously evaluates reality against them and acts to close the gap. Kubernetes runs that loop on infrastructure; SIE can run it on anything you can define.
Products
Three products, one engine.
Definition Manager
The authoritative API, lifecycle enforcer, and store of every governed GSM Definition — the heart of the engine.
Features
Operator
The GSM Definition runtime — for evaluating and executing GSM Definitions such as Norms and Mechanisms.
Features
Definition Blackboard Manager
The shared board where humans, AI, and tools think together about what your systems are — partial views composed into definition proposals.
Features