Rationalized Advanced Intelligence System Engine. Sovereign Infrastructure and "Local-First" Architecture for Critical Engineering.
Note
Completion of Initial Conception Phase (v0): This repository marks the culmination of six months of R&D focused on the RAISE core infrastructure and its foundational cognitive block.
- Repository Status: This repository is now archived as the historical "Open Core" foundation and is maintained in read-only mode for reference.
- v1 Transition: Active development has transitioned to a new dedicated repository for v1. All future features and roadmap updates will be managed there.
- Intellectual Property: In accordance with our licensing model, only the generated framework code is open-source. SysML schemas, internal databases, and specialized cognitive blocks remain proprietary intellectual property.
Important
π UPDATE v0.3.0 (Architecture Shift): RAISE has solidified its foundation as an Open Core Framework. This repository contains the highly secure infrastructure chassis (Rust, Wasm, Tauri). Cognitive intelligence modules, multi-agent solvers, and proprietary business logic are designed to be plugged into this public core via strict, isolated interfaces.
In critical engineering (Aerospace, Defense, Energy), sovereignty, data integrity, and strict decoupling are non-negotiable. RAISE provides a modern infrastructure aligning with Industry 4.0 standards (MBSE 2.0, SysML v2, RAMI 4.0) through a 100% Air-Gap capable architecture.
The public framework acts as an impenetrable shield and orchestrator, ready to host specialized cognitive engines without compromising the host system.
The following diagram illustrates the strict decoupling between the public infrastructure (Open Core) and the isolated abstraction layers designed to host proprietary cognitive engines.
flowchart TB
subgraph ClientLayer ["π₯οΈ Client Interfaces (Multi-Target Deployments)"]
direction LR
CLI["raise-cli<br/>(DevSecOps / CI Automation)"]
UI["raise-desktop<br/>(Tauri Secure Native UI)"]
EDGE["raise-edge<br/>(Embedded / Tactical Systems)"]
end
subgraph CoreLayer ["π‘οΈ raise-core (Rust Immutable Infrastructure)"]
direction TB
KERNEL["Micro-Kernel<br/>(Async Orchestrator)"]
DB["json_db<br/>(Handle-Based Integrity)"]
MBSE["model_engine<br/>(SysML v2 / Arcadia)"]
RULES["rules_engine<br/>(Deterministic Logic)"]
KERNEL <--> DB
KERNEL <--> MBSE
KERNEL <--> RULES
end
subgraph IsolationLayer ["π§© Abstraction & Sandboxing Layer"]
direction TB
WASM["Wasm Runtime<br/>(Isolated Plugin Execution)"]
STUBS["Cognitive Stubs<br/>(ai, agents, genetics APIs)"]
WASM <--> STUBS
end
subgraph PrivateIP ["π Proprietary Business Logic (Air-Gapped)"]
COGNITIVE["Private Cognitive Block<br/>(Enterprise IP, LLM, Rulesets)"]
end
%% Routing
CLI --> KERNEL
UI --> KERNEL
EDGE --> KERNEL
KERNEL <--> WASM
%% Strict API Contract boundary
STUBS -. "Strict Schema Contracts" .-> COGNITIVE
%% Styling
classDef core fill:#003366,stroke:#00AEEF,stroke-width:2px,color:#fff;
classDef client fill:#1e293b,stroke:#64748b,stroke-width:1px,color:#fff;
classDef isolate fill:#0f172a,stroke:#f59e0b,stroke-width:2px,stroke-dasharray: 5 5,color:#fff;
classDef privateIP fill:#7f1d1d,stroke:#ef4444,stroke-width:2px,color:#fff;
class KERNEL,DB,MBSE,RULES core;
class CLI,UI,EDGE client;
class WASM,STUBS isolate;
class COGNITIVE privateIP;
RAISE abandons fragile internal identifiers.
- Deterministic Relationships: The internal
json_dbengine strictly utilizes explicit Handles instead of volatile_idor auto-generated UUIDs. - Strict Data Agents: Data agents operate exclusively with explicit schema arguments to guarantee end-to-end data validation.
The engine is built on Rust to ensure absolute memory safety and high-performance topological simulation.
- Production Purity: RAISE enforces a strict zero-tolerance policy for testing artifacts. Absolutely no test conditions are included in production code. The runtime is pure, deterministic, and industrial-grade.
- Asynchronous Micro-Kernel: Efficient orchestrator built to handle non-blocking messages across highly decoupled modules.
How do we execute proprietary, critical AI models safely?
- Isolated Plugins: Extensibility is managed through WebAssembly (Wasm) sandboxes. This protects the core runtime from plugin panics or memory leaks.
- Cognitive Stubs: The
src/directory includes abstract interfaces (ai,agents,genetics,blockchain). These are pure structural contracts (APIs/Traits) waiting to be implemented by private, domain-specific intellectual property.
raise-desktop: A secure, lightweight, native user interface powered by Tauri.raise-cli: Designed for CI/CD automation. Smart Fallback: By default, if the--domainor--dbarguments are not explicitly specified during a command, the CLI automatically falls back to thecurrent_domainandcurrent_dbof the active session.
raise/
βββ raise-cli/ # DevSecOps command-line automation
βββ raise-desktop/ # Tauri-based secure local interface
βββ raise-edge/ # Lightweight deployments for constrained environments
βββ raise-core/ # The Core Infrastructure
βββ src/
βββ kernel/ # Core orchestrator
βββ json_db/ # Handle-based local data store
βββ rules_engine/ # Deterministic logical validation
βββ model_engine/ # MBSE 2.0 / SysML v2 data structures
βββ plugins/ # WebAssembly isolation layer
βββ code_generator/ # Ast-to-Source code weaving
βββ [ai/ agents/] # API Stubs for future cognitive extensions
- Local-First / Offline-First: Zero dependency on external SaaS or Cloud APIs. The entire infrastructure is designed to operate in highly classified, completely disconnected zones.
- Zero-Trust Ready: Verification at every step, enforced by the schema-driven data agents and immutable state management.
git clone https://github.com/Condorcet-Continuum/raise.git
cd raise
cd crates/raise-cli
cargo build --release
cd crates/raise-desktop
npm install
cargo tauri dev