Cloud platform engineer focused on AWS and Azure: Kubernetes platforms, GitOps, IaC, and cloud security. Currently a DevOps Engineer at Snorkel AI.
I build everything here twice where it counts, once on AWS and once on Azure, to prove the pattern rather than the provider.
| Project | What it shows |
|---|---|
| aws-developer-platform | Internal Developer Platform on EKS: ArgoCD, Crossplane self-service infra, Kyverno guardrails, Backstage golden path |
| azure-developer-platform | The same paved-road pattern on AKS with a deliberately different toolchain: Flux, Crossplane, Workload Identity |
| cloud-security-lab | Full MITRE ATT&CK kill chain against an intentionally vulnerable AWS environment, with Falco + OPA Gatekeeper detection |
| zero-trust-identity-pipeline | Entra ID Conditional Access, PIM, and Sentinel automation, all in Terraform |
| aws-cost-intelligence-dashboard | FinOps platform: z-score anomaly detection and spend forecasting on Cost Explorer data |
| nfl-data-reliability-platform | SRE-style ingestion on Azure Container Apps with Prometheus SLIs, Grafana burn rates, and runbooks |
| llm-gateway | Multi-provider LLM proxy on ECS Fargate with caching, observability, and LLM-as-judge evals |
Every project is Terraform-managed, deployed with GitHub Actions OIDC (no stored cloud credentials), and torn down after demos to keep spend near zero.
- Platforms over pipelines: self-service golden paths, policy-as-code guardrails, GitOps reconciliation
- Security by default: least-privilege IAM, managed identities, encryption at rest, no static credentials anywhere
- Cost-aware: FinOps practices applied to my own lab spend
- Portfolio: jordandesigns.io
- LinkedIn: jordan-nelson