CLOUD ARCHITECT LAB
Architecture Case Studies
Practical cloud architectures developed through a repeatable engineering lifecycle. Each case study connects requirements, architecture, implementation, validation, and lessons learned.
CAL-005
Deterministic Mermaid Evidence Extraction
Evidence as Data.
Evaluation by Evidence.
Mermaid • Deterministic Parsing • Controlled Rendering • AI Evaluation
Preserves architecture evidence through deterministic Mermaid extraction, then tests whether multiple AI evaluators can detect explicit defects and missing required relationships.
View Case Study →CAL-004
AI Architecture Evaluator
Architecture as Code.
Evaluation by Evidence.
Mermaid • Governed Knowledge • Structured Findings • AI Evaluation
Tests grounded architecture conformance evaluation: whether AI can inspect machine-readable architecture against authoritative engineering requirements and support its findings with evidence and citations.
View Case Study →CAL-003
AI Knowledge Assistant
Knowledge as Code.
Authority by Design.
Amazon Bedrock • Titan Embeddings V2 • S3 Vectors • Claude Haiku 4.5
Implements a governed Retrieval-Augmented Generation architecture with metadata-aware retrieval, grounded responses, citations, adversarial evaluation, and deterministic knowledge-authority controls.
View Case Study →CAL-002
Enterprise VPC Peering
Infrastructure as Code.
Architecture as Code.
AWS • Terraform • VPC Peering • NAT Gateway • Network ACLs
Extends the CAL networking foundation into a two-VPC enterprise architecture with reusable modules, private management paths, independent outbound egress, and defense-in-depth controls.
View Case Study →CAL-001
AWS Networking Foundations
Infrastructure as Code.
Architecture as Code.
AWS • Terraform • Amazon VPC • Mermaid
Establishes the reusable networking foundation for Cloud Architect Lab through custom VPC design, public and private subnet separation, explicit routing, security controls, and validation.
View Case Study →More Than Working Code
Cloud architecture is not complete when resources deploy successfully. Each case study is expected to explain what the system must accomplish, why the design was selected, how it was implemented, how it was validated, and what should improve next.