I sit at the intersection of two roles companies are paying a premium for in 2026: the engineer who can deploy AI systems reliably, and the engineer who can use AI to operate infrastructure intelligently.
On the DevOps side: I led an EKS migration at Malomatia that cut infrastructure costs by 20% — earning Star of the Month. I've standardised CI/CD pipelines across 3 engineering teams, shipped GitOps with ArgoCD, and maintained production Kubernetes clusters with Terraform, Prometheus, and Grafana.
On the Agentic AI side: I architected KubeSentinel, an autonomous AI-SRE that cut MTTR by 80%+ using LangGraph, RAG, and MCP tool-calling for GitOps PR generation. I then led a team of 4 building the Agentic NOC Platform — 33-server government deployment achieving sub-60-second detection-to-remediation.
What's next: deeper work on multi-agent orchestration, MCP server design, and agentic workflows in regulated environments. The thesis I'm building toward: the best SREs in 2027 will be agents, and the engineers who built them will be invaluable.