I study how security systems fail: trust boundaries erode, identity paths compromise, and governance gaps widen before anyone notices.
Currently building Building akintoye.ca on AWS with Terraform: Terraform modules for this site on AWS with private S3, CloudFront OAC, and WAF.
Akíntóyè O.
Infrastructure and observability work with corroborating artifacts
Production infrastructure for the Akintoye platform as a Next.js static export at akintoye.ca. Six Terraform modules wire Route53, ACM, private S3, CloudFront, WAF, and CloudWatch with content deploys isolated from infrequent IaC applies.
Production monitoring for akintoye.ca combining AWS CloudWatch edge metrics with homelab deep observability. Prometheus, Grafana, Loki, and Alertmanager on the platform node deliver external availability probes, SSL expiry tracking, deployment correlation, and infrastructure health dashboards.
A two-node homelab that separates compute from platform services. The HP Ryzen 3 laptop runs Prometheus, Grafana, Loki, and Tailscale subnet routing 24/7 while the Mac M2 handles Docker Desktop development stacks, preserving monitoring during compute maintenance.
Original models for trust erosion and identity compromise
Modern identity systems create chains of trust that, when compromised at any point, enable cascading access escalation across organizational boundaries.
Trust boundaries in complex systems do not fail catastrophically - they erode gradually through accumulated exceptions, implicit dependencies, and architectural drift until the boundary exists only in documentation.
Observations from production systems and governance gaps
SaaS proliferation creates credential reuse patterns that bypass organizational security controls entirely.
Push-based MFA creates a novel social engineering surface that exploits human tolerance for system interruptions.
Cloud IAM governance is failing because organizations apply legacy governance frameworks to fundamentally different identity architectures.
Identity-based attacks exploit a fundamental detection asymmetry: each action appears individually legitimate.
Long-form analysis on systems failure and security culture
Security awareness programs fail not because people are careless, but because they treat a systems problem as an individual knowledge deficit.
Google's forced password resets reveal a deeper truth: identity risk is strategic, not operational, and leadership must own the perimeter.
The first confirmed AI-orchestrated cyber-espionage campaign has been detected. Cybersecurity has officially entered its autonomous era.
After falling sick and pushing through an exam anyway, failure became the catalyst for a deeper lesson in leadership, resilience, and growth.