ModulesSAI-120
SAI-120 table of contents
Sovereign AI Reference Architecture
Design vendor-neutral architecture layers, deployment patterns, assurance points, and decision records.
Learning outcomes
What you should be able to do
- Select an appropriate deployment pattern
- Document portability and concentration decisions
- Place assurance points across the lifecycle
Curriculum
Work through 5 sections in order.
The chapters are individually addressable documentation pages. You can link directly to a concept from another program, architecture decision, or implementation guide.
Architecture foundations
Connect system purpose and sovereignty dimensions to explicit architecture requirements.
Introduction to sovereign AI architecture
Translate system purpose, sovereignty requirements, threats, and operating constraints into a vendor-neutral architecture method.
Sovereignty control dimensions
Evaluate data, model, infrastructure, operational, and evidence control as separate but connected dimensions.
Define the AI system
Set the intended purpose, actors, assets, dependencies, lifecycle, and system boundary before selecting controls.
Layers, views, and interfaces
Separate responsibilities and describe the system consistently across logical, deployment, trust, lifecycle, and evidence views.
Sovereign AI reference-architecture layers
Separate experience, workload, knowledge, model, gateway, platform, supply-chain, operations, and evidence responsibilities.
Architecture views and controlled interfaces
Use context, logical, deployment, trust, lifecycle, operational, and evidence views with consistent identifiers and contracts.
Trust boundaries and data flows
Map where trust changes and where data, artifacts, requests, identities, tools, and evidence cross a boundary.
AI assets and attack surfaces
Inventory the data, models, prompts, artifacts, identities, tools, interfaces, infrastructure, and evidence that require protection.
Patterns, decisions, and exit
Select deployment patterns, record trade-offs, and preserve the practical ability to change direction.
Deployment boundaries and patterns
Compare public cloud, private cloud, on-premises, hybrid, edge, restricted-network, and air-gapped patterns without treating one as automatically sovereign.
Architecture decision records
Record context, options, control consequences, evidence, dependencies, limitations, and reversal triggers for material decisions.
Portability, reversibility, and exit
Design and test the ability to change models, runtimes, data stores, infrastructure, providers, and operating arrangements.
Control and assurance
Place control objectives and evidence at the architecture points where material decisions are enforced.
Write effective control objectives
Translate sovereignty goals into testable statements of what must be allowed, prevented, approved, observed, retained, and recovered.
Evidence by design
Design evidence alongside controls so important decisions, releases, configurations, and operating events can be verified.
Apply and assess
Produce a defensible reference architecture and test its completeness and traceability.
Reference-architecture workshop
Create and defend a multi-view sovereign AI architecture, decision set, control map, evidence map, and readiness backlog.
SAI-120 knowledge check
Verify architecture completeness, traceability, boundary control, decision quality, portability, operability, and evidence coverage.
Practical completion package
- System context and architecture-view set
- Layer and component responsibility map
- Trust-boundary and controlled-interface map
- Architecture decision-record set
- Control, evidence, portability, and readiness plan
Current release boundary
This public curriculum provides a vendor-neutral architecture method and review structure. It is not a production design approval, vendor certification, or substitute for workload-specific engineering, security, legal, and operational review.