01
BANKING & FINANCE
APRA-Compliant Banking AI Stack
Governance and security controls for APRA-regulated AI, spanning model accountability, information security and operational resilience.
12–16 weeksAPRA CPS230APRA CPS234
Start from a control sequence, not a template card.
Six enterprise starting points that show how controls connect across a system. Use the sequence, inspect each control, then adapt the architecture to the estate you actually operate.
02
GOVERNMENT
AU Government AI Governance Stack
Governance-first architecture for public-sector AI with risk classification, ethics review, approval, human oversight and traceability.
16–20 weeksAU Government AI EthicsPrivacy ActISO 42001
03
CROSS-INDUSTRY
Enterprise RAG + Knowledge Platform
A governed retrieval platform combining grounding, evaluation, lineage, security and production observability.
8–12 weeksPrivacy ActGDPR
04
TECHNOLOGY
Production Agentic AI Stack
A controlled agent runtime spanning orchestration, tool execution, sandboxing, durable workflow state and telemetry.
10–14 weeksEU AI Act Art. 9ISO 42001
05
HEALTHCARE
Healthcare AI Safety Stack
Safety-first architecture for clinical and healthcare contexts with mandatory oversight, escalation and reconstructable decisions.
14–18 weeksPrivacy ActTGA contextISO 42001
06
CROSS-INDUSTRY
Multi-Agent Enterprise Automation Stack
An enterprise automation control path covering routing, multi-agent coordination, durable state, observability and audit evidence.
12–16 weeksISO 42001NIST AI RMF
ADAPT, DO NOT COPY
Reference architecture is the opening move.
Run your actual architecture through EAAPL to find which controls apply, where trust boundaries differ and what evidence leadership will need.
Analyse the real system →