I am building an Agentic AI practice on top of 5+ years of cloud platform work. I already run LLM integration in production: an AI gateway, LLM observability, MCP-based automation. My view is simple. AI should meet the same security and governance bar as everything else we run.
LLM Integration
Bedrock and Claude in production, with MCP agent skills and LLM observability.Secure by Design
Production platforms with security, auditability and cost control built in.Cloud & Delivery
EKS on AWS with Terraform, and the same artefact from commit to production.Professional Experience
I designed and ran secure cloud-native platforms on AWS, EKS and Terraform, then built the AI layer on top: an AI gateway, LLM observability and agentic automation. The point was to give business teams AI that meets the same security and governance bar as everything else.
- Brought AI to business teams by integrating AWS Bedrock and Claude behind a central AI gateway (LiteLLM), giving unified access, model routing and cost governance
- Built monitoring and traceability over LLM traffic to meet security audit and DLP requirements, shipping logs to CloudWatch and S3
- Developing AI agent skills to move SOP-based workflows towards intelligent automation
- Built and deployed secure, scalable platforms for DevOps and business teams on Kubernetes (EKS, Karpenter, Cilium), with Terraform across AWS
- Deployed and maintained EUC (End-User Computing) solutions: ready-to-use environments for developers and data scientists, via Coder
- Automated CI/CD pipelines (GitLab CI, GitHub Actions) for app and infrastructure delivery, backup/restore, and platform upgrades
- Shared knowledge and practice around AI workflows and agentic paradigms (n8n, MCP); team SME for observability platform selection
As a DevOps Specialist, I migrated the DCP (Digital Cooking Platform) from EC2 to Kubernetes (EKS) using Terraform and Ansible, and built a Kubernetes integration platform that gave teams Jenkins instances on demand.
- Implemented the target cloud architecture and migrated the Digital Cooking Platform from EC2 to Kubernetes (EKS) with Terraform automation
- Provisioned on-demand Jenkins instances for teams through a GitOps workflow with ArgoCD
- Created Grafana dashboards and Prometheus alerts for observability
- Optimised Kubernetes resources to improve performance and reduce costs (FinOps)
- Deployed the CrowdStrike antivirus solution across the Kubernetes clusters
- Created and migrated Jenkins jobs; configured GitHub Actions workflows for CI/CD automation
- Used IaC with Ansible and Terraform to automate infrastructure component deployments
On assignment for SNCF as a DevOps Engineer in the Ops team on TRAIN CONNECT, I modernised and kept in service the real-time ground-to-train data exchange platform behind connected trains. A large-scale hybrid AWS and on-premises system spanning several teams.
- Orchestrated and ran the containerised environments (Docker, AWS ECS and EKS) that keep the real-time data flows scalable
- Developed Helm charts and automated deployments via Jenkins and Ansible
- Designed and deployed highly available, secure AWS architectures with Terraform, automating IAM, security groups and certificates
- Implemented cross-cutting platform components (Traefik, HiveMQ, Nginx)
- Load testing, resilience analysis and monitoring (Datadog, New Relic), with improvement proposals
- Designed and optimised job performance; raised and tracked tickets with AWS support
- Collaborated with development and testing teams to resolve anomalies and write technical documentation
DevOps consultant and Fullstack Developer at Cromology Lyon, where I led the client's DevOps transformation and CI/CD automation while contributing to fullstack feature development.
- Led client-facing DevOps transformation: CI/CD pipelines with GitLab CI targeting on-premises RHEL
- Automated packaging and delivery processes with Azure DevOps
- Designed and developed backend PHP and frontend React modules for a quote simulator
- Wrote technical and application architecture documentation
- Active participation in the DevOps community and tech watch around microservices architectures
- Harmonised DevOps practices and automated CI/CD pipelines (SVN → GitLab migration)
- Benchmarked cloud solutions (OpenStack, Azure, OpenShift) and led Docker/Kubernetes acculturation
- Planned and documented cloud migration strategies
- Developed C#/.NET 4.6 applications; led the technical foundation team and performed code quality audits