Security, Cost & Well-Architected Frameworks
Compliance & Governance
Enterprise architects must understand compliance requirements and governance frameworks. This knowledge demonstrates business acumen alongside technical expertise.
Compliance Frameworks
Common Standards
| Standard | Industry | Key Requirements |
|---|---|---|
| SOC 2 | SaaS/Tech | Security, availability, confidentiality |
| HIPAA | Healthcare | PHI protection, audit trails |
| PCI DSS | Finance | Cardholder data security |
| GDPR | EU Data | Data privacy, consent, right to erasure |
| FedRAMP | US Government | Security controls, continuous monitoring |
| ISO 27001 | Global | Information security management |
AWS Compliance Programs
AWS maintains compliance certifications that customers can inherit:
| Certification | Benefit | Customer Responsibility |
|---|---|---|
| SOC 1/2/3 | Financial controls audited | Application-level controls |
| HIPAA | BAA available | PHI handling policies |
| PCI DSS Level 1 | Infrastructure compliant | Application PCI controls |
| ISO 27001 | Security framework | Implement controls |
| FedRAMP High | Government approved | Customer-side controls |
Interview Question: HIPAA Architecture
Q: "Design a HIPAA-compliant architecture for a healthcare application."
A: Layered compliance approach:
HIPAA Architecture
├── Network Layer
│ ├── Private subnets for PHI
│ ├── VPN/Direct Connect only
│ └── No public internet access for PHI
├── Data Layer
│ ├── Encryption at rest (KMS)
│ ├── Encryption in transit (TLS)
│ └── Field-level encryption for SSN, DOB
├── Access Layer
│ ├── Role-based access (IAM)
│ ├── MFA required
│ └── Break-glass procedures logged
├── Audit Layer
│ ├── CloudTrail (all regions)
│ ├── VPC Flow Logs
│ └── Application audit logs
└── Administrative
├── BAA with AWS signed
├── Employee training documented
└── Incident response plan
AWS Governance Tools
AWS Organizations
Structure for multi-account governance:
Organization Root
├── Core OU
│ ├── Security Account
│ ├── Logging Account
│ └── Network Account
├── Production OU
│ ├── Prod Account A
│ └── Prod Account B
├── Non-Production OU
│ ├── Dev Account
│ └── Staging Account
└── Sandbox OU
└── Sandbox Account
Service Control Policies (SCPs)
| SCP Type | Purpose | Example |
|---|---|---|
| Deny List | Block specific actions | Deny EC2 in non-approved regions |
| Allow List | Permit only specific | Only allow approved services |
| Guardrails | Prevent risky changes | Prevent CloudTrail disable |
Example: Region Restriction SCP
{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Deny",
"Action": "*",
"Resource": "*",
"Condition": {
"StringNotEquals": {
"aws:RequestedRegion": [
"us-east-1",
"us-west-2",
"eu-west-1"
]
}
}
}
]
}
AWS Config
| Feature | Use Case |
|---|---|
| Configuration Recording | Track all resource changes |
| Conformance Packs | Pre-built compliance rules |
| Remediation | Auto-fix non-compliant resources |
| Aggregation | Multi-account compliance view |
Interview Question: Governance at Scale
Q: "How would you implement governance for 100+ AWS accounts?"
A: Centralized governance model:
-
AWS Organizations
- OU structure by environment/business unit
- SCPs for guardrails
- Consolidated billing
-
Control Tower
- Landing zone automation
- Account factory for provisioning
- Guardrails (preventive + detective)
-
Centralized Services
- Security Hub aggregation
- Config aggregator
- CloudTrail organization trail
-
Automation
- Account vending machine
- Baseline CloudFormation StackSets
- Automated compliance reporting
Data Governance
Data Classification
| Classification | Examples | Controls |
|---|---|---|
| Public | Marketing materials | Minimal |
| Internal | Employee data | Encryption, access control |
| Confidential | Customer PII | Encryption, audit, DLP |
| Restricted | Payment data, PHI | All above + additional |
Data Residency
| Requirement | Implementation |
|---|---|
| Region restriction | SCPs, AWS Config rules |
| Cross-border transfer | Legal review, encryption |
| Data localization | Single-region deployment |
| Backup location | Same-region backup policies |
Data Lifecycle
Creation → Classification → Storage → Use → Archival → Deletion
↓ ↓ ↓ ↓ ↓ ↓
Tagging Auto-classify Tiering Access Glacier Secure
(Macie) (S3) Logs (S3) Delete
Audit and Reporting
Audit Trail Requirements
| Log Type | Service | Retention |
|---|---|---|
| API Calls | CloudTrail | 7 years (compliance) |
| Network | VPC Flow Logs | 1 year |
| Application | CloudWatch Logs | Per requirement |
| Database | RDS/Aurora logs | Per requirement |
Compliance Reporting
| Tool | Function |
|---|---|
| AWS Artifact | Download compliance reports |
| Security Hub | Compliance score dashboard |
| Config | Compliance timeline |
| Audit Manager | Automated evidence collection |
Interview Question: Audit Readiness
Q: "How do you prepare for a SOC 2 audit?"
A: Continuous compliance approach:
-
Evidence Collection (Audit Manager)
- Automated evidence gathering
- Control mapping to SOC 2
-
Continuous Monitoring
- Config rules for each control
- Security Hub for posture
- Automated alerting on drift
-
Documentation
- Architecture diagrams current
- Runbooks documented
- Change management logs
-
Pre-Audit
- Internal assessment
- Gap remediation
- Mock audit walkthrough
Cloud Governance Best Practices
Governance Framework
| Component | Implementation |
|---|---|
| Policies | Written, versioned, accessible |
| Standards | Technical requirements documented |
| Procedures | Step-by-step runbooks |
| Guidelines | Best practices, recommendations |
Change Management
Request → Review → Approve → Implement → Verify → Close
↓ ↓ ↓ ↓ ↓ ↓
Ticket CAB or Workflow CI/CD Testing Audit
Auto- (SNS) Pipeline Checks Log
approve
Key Insight: Compliance is not a one-time achievement—it's continuous. Automate compliance checks, evidence collection, and remediation to maintain audit readiness at scale.
Next module: Behavioral and negotiation strategies for architect interviews. :::