I build production AI systems — and I run them with the same reliability

discipline I've spent a decade applying to cloud infrastructure. AI is where

most of my work lives now; DevOps is the backbone that keeps it dependable.

On the AI side, I design and operate LLM and RAG platforms end to end:

retrieval pipelines, fine-tuning and inference services (vLLM, MLflow,

JupyterHub), and production integrations on top of OpenAI and Anthropic

models. I treat AI workloads as first-class production systems — they get

the same SLOs, observability, cost controls, and GitOps delivery as anything

else, instead of living as fragile notebooks that break the moment they meet

real traffic.

I also apply AI back into DevOps itself: agentic automation and LLM-assisted

workflows for code review, infrastructure-as-code generation, runbook and

incident triage, and faster root-cause analysis — automation that augments

the platform team rather than adding noise.

That all stands on 10+ years of cloud and platform engineering. I've

architected infrastructure, automated CI/CD, and scaled Kubernetes workloads

across AWS and GCP for companies in aviation, insurance, manufacturing, and

software — including Allianz Technology, Aer Lingus, Molex, and Yields.

My focus is the part that decides whether a platform survives production:

reliable delivery, sane IAM, observability that catches problems before users

do, and infrastructure your own engineers can confidently maintain.

What I work with:

• AI / LLMOps — RAG pipelines, fine-tuning & inference (vLLM, MLflow),

  OpenAI & Anthropic in production, agentic automation

• Kubernetes platforms — Helm, ArgoCD/Flux, GitOps delivery, multi-cluster

• Infrastructure as Code — Terraform (HashiCorp Certified), Ansible, Pulumi,

  Crossplane

• Cloud — AWS (EKS, RDS, IAM, VPC) and GCP (GKE, Cloud SQL), multi-cloud

  landing zones, disaster recovery

• CI/CD & Observability — GitLab CI, Jenkins, Prometheus, Grafana, Loki,

  OpenTelemetry

I'm the Director of DigitalCloud Ops Ltd, a UK consultancy building

production-grade Kubernetes platforms, GitOps pipelines, and AI systems for

teams that can't afford to guess — "engineering reliability into modern

infrastructure." I also build Brevyn, a Claude Certified Architect exam-prep

platform.

If you're taking an LLM prototype to production, standing up a Kubernetes

platform, or want AI woven into your delivery and operations the right way —

let's talk.

? Bucharest / London · Open to remote consulting worldwide