Architecture
Best-Practice Library

Core principles and philosophies guiding all design decisions across the practice.

Canonical structural, integration, security, and data patterns.

Common design mistakes and legacy pitfalls to recognise and avoid.

ADRs documenting rationale, context, alternatives, and impact of key decisions.

Perspectives: logical, physical, process, deployment, and scenario.

End-to-end reference scenarios and trade-off analysis for scalable systems.

Component and service-level design patterns and practices.

Context mapping, bounded-context patterns, and aggregate design.

TOGAF, NIST CSF, ISO 27001, Zachman, and Gartner — mapped to practice.

Frontend, backend, data, cloud, DevOps, practice circles, and engagement models — how we deliver across the stack.

Multi-cloud reference architectures across AWS, Azure, and GCP.

IaC, networking, CI/CD, and operational excellence practices.

Data modeling, governance, lineage, mesh patterns, and analytics.

API, event, messaging, workflow, and partner integration patterns.

End-to-end security design: AuthN/AuthZ, encryption, and cloud controls.

Standards mappings and compliance controls for enterprise environments.

How the governance system itself runs: checklists, review templates, roles and authority, and scorecards measuring whether governance is working.

Performance, availability, scalability, and security NFR evaluation.

The foundational observability primitives — incident response, structured logging, metrics, SLIs/SLOs, and distributed tracing — that any production system depends on, separate from AI-specific monitoring.

The artefacts a responder picks up under pressure — incident, migration, and rollback runbooks — as designed documents with structure, calibration, and lifecycle, distinct from the operational disciplines they execute.

The artefacts a reviewer applies during structured review work — architecture, deployment readiness, and security checklists — as designed coverage instruments with dimensions, calibration, and tier-awareness, distinct from the governance disciplines that produce them.

The engineering tools the team uses to build, validate, and operate systems — AI agents, CLI tools, scripts, and validators — as designed artefacts with composability, observability, and lifecycle properties, distinct from the systems they help build.

Production-grade AI systems: architecture, responsible-AI engineering, observability, retrieval, and the AI-specific threat surface.

AI-native patterns, LLMOps, RAG, agentic systems, and governance.

Capability scoring and maturity criteria for engineering advancement.

Scoring templates for continuous architecture quality measurement.

AI readiness, modernization roadmaps, and foundational strategy.

Platform evolution and capability uplift roadmaps.

Step-by-step playbooks for common architecture challenges.

The structured artefacts the team uses to record decisions, document designs for review, and rate services across dimensions — ADRs, review templates, and scorecards — as designed instruments whose fields, lifecycle, and calibration determine whether they accumulate as institutional memory or as ceremonial paperwork.