Abdulraheem Shonubi · Senior Cloud & DevOps Engineer · AI-Assisted Platform Engineering

Production should be boring. I build the platforms that keep it that way.

Five years running production on Azure and AWS for fintech and AI-powered products. Everything in code, every change reviewed, no surprises on the bill. AI agents do the legwork under least privilege; design, security and production decisions stay with me.

Selected work

Case studies from production: the problem, the constraints, the decisions and what changed.

  • 13 min read

    Private isn't segmented: the PCI DSS Requirement 1 gaps I keep finding in the cloud

    Cloud card data environments often look segmented on a diagram: no public IPs, a firewall in front. These are the seven PCI DSS Requirement 1 gaps I keep finding behind that, how I close them, and the guardrails that stop them coming back.

    Recurring Requirement 1 gap patterns
    7
    Gaps closed on my latest assessment
    All
    Open ports without approved need
    0

    Stack: PCI DSS v4.0 · Azure · Azure Firewall · Network Security Groups · Microsoft Defender for Cloud · Azure CLI

  • 11 min read

    Reviving drifted production Terraform: 40% to 100% under code, zero changes

    A production Azure estate's Terraform had silently stopped matching reality, and its main branch would have destroyed or downgraded live resources. I led a human-governed, AI-assisted revival that brought every resource under Terraform with 135 imports and zero changes to Azure.

    Resources managed by Terraform
    100%
    Before: 40%
    Harmful plans on main branch
    0
    Before: 5
    Resources imported
    135

    Stack: Terraform · Azure · GitHub Actions · AKS · Azure Firewall · Claude Code

  • 7 min read

    Azure tag compliance from 20% to 100%, and the audit it unlocked

    Only 20% of a production Azure estate was tagged, so nobody could say who owned what or what it cost. I standardised tags with Azure Policy to reach 100% of taggable resources, then used them to find orphaned resources and feed a 30% cost reduction.

    Tag compliance (taggable resources)
    100%
    Before: 20%
    Azure cost, wider cost programme
    −30%
    Tagging decisions needed
    31 groups
    Before: 232 resources

    Stack: Azure · Azure Policy · Azure CLI · Azure Cost Management · Excel

All case studies →

Experience

Where I've owned production, and what changed because of it. The résumé has the detail.

  1. Senior DevOps Engineer, Sparkle Nigeria

    Jun 2026 – PresentRemote, NigeriaFintech, payments

    Own the Azure platform behind customer-facing payment services: Kubernetes, infrastructure as code, CI/CD, observability and PCI DSS network controls.

    • Every production Kubernetes upgrade and release delivered with zero downtime.
    • PCI DSS Requirement 1 owned end to end for the card data environment, with clean quarterly ASV scans.
    • GitOps with ArgoCD and reusable Helm charts, so payment services ship to production the same reviewed, auditable way.
  2. Senior DevOps & Cloud Operations Engineer, GameIT

    Jul 2026 – PresentPart-timeRemote, KSAAI-powered EdTech

    Own the Azure platform behind an AI-powered games product and its ML services, and its scaling and FinOps roadmap.

    • Introduced Terraform, so every environment is reproducible, reviewed infrastructure as code.
    • Designed least-privilege, fail-closed guardrails that let AI coding agents work on production infrastructure safely.
    • Removed long-lived secrets from CI/CD by moving every pipeline to workload identity federation.
  3. Senior Technical Release Manager, Spacepointe Inc

    May 2026 – Oct 2026Remote, United StatesCommerce platform

    Release authority for a web, API and Android platform: 12 engineers, 3 parallel streams, five environments from Dev to Prod.

    • Authored the Release Management Policy and six-gate SDLC, enforced through a Change Advisory Board.
    • Drove a 9-point release-readiness plan and ended a recurring production instability in the pipeline.
  4. DevOps / Cloud Infrastructure Engineer, Teknowledge

    Feb 2023 – Mar 2026Remote, Nigeria

    Built and ran production infrastructure on Azure: migration, infrastructure as code, GitOps and observability.

    • 30% lower hosting costs from a production migration to Azure with zero unplanned downtime.
    • 99.99% uptime across 50+ Linux and Windows workloads on GitOps-delivered AKS.
    • Incident detection ~60% faster and response cut from 45 to under 15 minutes.
  5. Junior SRE / DevOps Engineer, Huawei Technologies

    Jan 2022 – Jan 2023Remote, NigeriaTelecoms

    Automation and monitoring for carrier-grade infrastructure serving major mobile network operators.

    • ~40% faster environment setup by automating Azure and AWS provisioning.
    • 20% fewer post-deployment incidents with GitHub Actions pipelines and Prometheus/Grafana monitoring.

Full résumé (PDF) →

How I work with AI agents

Human-governed, AI-assisted platform engineering: one engineer covering more ground without lowering the bar.

Least privilege, for agents too
Agents get only the access the task needs: read and plan, never apply. A pre-tool-use hook blocks apply, destroy, state edits and direct cloud CLI or API calls, and fails closed. The guard rules are tested before an agent touches a live environment.
I own the decisions
Architecture, security, cost and what fits the systems already running. Every plan is reviewed, every drift is decided by a person, and I run every production change.
Agents do the legwork
Investigation, inventories, first drafts of Terraform and pipelines, plan summaries, runbooks and decision logs. The groundwork that usually slows a small team down.
Everything leaves a paper trail
Work is planned, tracked and documented in the repo as it happens, so every accepted change and exception can be audited later.

See it in production: 135 Terraform imports, zero changes to Azure →

Certifications

Verified credentials, each linked to its issuer.

Labs & open source

Reference builds I publish to explore and teach a pattern end to end. Each has its code and a write-up.

All labs & open source →

Skills

The tools I use in production. The résumé has where and how.

Cloud
Microsoft Azure, AWS
Containers & orchestration
Kubernetes (AKS, EKS), Docker, Helm, ArgoCD, Kustomize
Infrastructure as code
Terraform, Bicep, ARM
CI/CD & release
GitHub Actions, Azure DevOps, GitOps, release management
Observability & SRE
Datadog, SigNoz, OpenTelemetry, Prometheus, Grafana, Azure Monitor, SLIs/SLOs, incident response
Security & compliance
PCI DSS (network segmentation, firewall and NSG rule governance, vulnerability remediation), Microsoft Entra ID, AWS IAM, secrets management, Microsoft Defender for Cloud
Networking
VNet/VPC, NSGs and security groups, load balancers, DNS, NGINX, Kubernetes ingress
Scripting & OS
Bash, PowerShell, Azure CLI, AWS CLI, Linux, Windows Server
AI-assisted engineering
Claude Code, agentic IaC workflows (investigation, drift triage, plan review), least-privilege agent access, fail-closed guardrail hooks, human-in-the-loop change control

Contact

Open to senior cloud, DevOps and platform engineering roles and contracts with remote-first teams.

raheem.shonubi@gmail.com