Cloud DevOps Engineer / SRE / Platform

AbdessamadeAbarchihi

I build secure, highly available systems across AWS, Azure and OCI, and automate the pipelines that ship them. Kubernetes at the core, infrastructure as code everywhere.

View work Resume

Available for work

Casablanca, Morocco · UTC+1

BuildAutomateScaleObserveSecureShip

01 / System / interactive

Ask the shell.

A real prompt. Try help, whoami, stack, uptime or deploy.

abdessamade@abarchihi · zsh

Last login: infrastructure that stays up.

abdessamade@abarchihi ~ % whoami

Abdessamade Abarchihi

Cloud DevOps Engineer · SRE · Platform Engineer

Casablanca, Morocco · working across time zones

Type `help` to list available commands.

02 / Strengths / how I work

What I bring.

The habits behind the systems, rather than one project’s numbers. These carry across stacks, clouds and teams.

01
Reliability first
Design for the failure case: health checks, rollback paths and alerting wired in before launch, not after the first incident.
02
Everything as code
Infrastructure, pipelines and policy live in version control, so any environment can be rebuilt from a clean checkout.
03
Kubernetes at depth
Cluster upgrades, networking, autoscaling and multi-tenant workloads, run in production, not just passed in an exam.
04
Security by default
Least-privilege access, secrets kept out of source, and scanning inside the pipeline instead of bolted on at review time.

03 / Selected work

All projects →

Things I shipped.

2025Platform / AI~/projects/kube-cortex

KubeCortex

View full size →

AI-driven, multi-cluster Kubernetes management and monitoring platform inspired by Rancher, with a LiteLLM + RAG diagnostic engine that reads logs and suggests automated fixes.

Kubernetes
FastAPI
LiteLLM
RAG
Helm

github.com/Abdessamade03View repo →

2025Agentic AI~/projects/smartgrocy

Agentic AI grocery planning engine deployed on Azure. Provisioned Container Apps infrastructure with Terraform, secured production with Private Links and Managed Identities, and traced flows with Application Insights.

Azure
Terraform
Gemini API
LangGraph

Client projectVisit site →

2026Kubernetes / LLM Serving~/projects/kuberay-inference

KubeRay Inference

View full size →

Self-hosted LLM inference on a Kubernetes cluster bootstrapped with kubeadm across a master and worker nodes. The KubeRay operator watches a RayService CRD and reconciles a RayCluster of head and worker pods, with Ray Serve deployments hosting the models and scaling replicas across the workers.