Started in structural engineering — load paths, safety factors, failure margins. Pivoted into software in 2023 and never lost the instinct: systems fail from unexamined assumptions, not bad code.
Today I build multi-tenant SaaS platforms and the Kubernetes/Terraform infrastructure underneath them, and I'm the founder of Future Eagle IoT. Currently deep in an AI/ML infrastructure transition — treating agents and RAG pipelines with the same rigor as any other production service.
flowchart LR
A[Civil Engineering] --> B[Software] --> C[DevOps] --> D[Cloud & Platform Engineering] --> E[AI Infrastructure]
style A fill:#2b3137,stroke:#4C8DFF,color:#fff
style E fill:#2b3137,stroke:#4C8DFF,color:#fff
| Project | Stack | What it does |
|---|---|---|
| AgentWarden |
Go · K8s admission webhooks | Sandboxes AI coding agents inside Kubernetes — a policy layer between "agent suggests" and "agent executes." |
| FlexTier |
Java · K3s · ECS Fargate | Migrates a legacy Java WAR two ways in parallel, with FinOps tagging to compare real cost trade-offs. |
| Blog-Platform-CloudOps |
Spring Boot · React · Terraform | Full-stack app with Docker, Jenkins, K8s, Prometheus & Grafana wired in from the first commit. |
| ticket-forge-devops |
Event-driven microservices | High-concurrency ticketing sandbox modeling flash-sale traffic patterns. |
| GitOps-on-k8s-using-ArgoCD |
Kubernetes · ArgoCD | Git as the single source of truth for cluster state — no kubectl apply from a laptop. |
Private / proprietary: Future Eagle Fleet (multi-tenant fleet SaaS — Spring Boot, PostgreSQL RLS, Keycloak) and Eagle IoT (industrial pipeline — PLC → ESP32 → Kafka → TimescaleDB).
Streak card and snake animation activate once profile-widgets.yml is enabled on this repo (included below).
Long-form notes on DevOps, DevSecOps, Cloud, MLOps & AIOps at the blog — including the Engineering AI and TicketForge book series, written alongside the projects above.


