Expertise

    DevSecOps & Kubernetes Security Engineer — CI/CD, GKE and Supply Chain

    Abraham O. Samuel builds security into delivery pipelines and Kubernetes platforms through SAST, DAST, container scanning, signed artifacts, policy-as-code, workload identity, network policy and runtime detection.

    Last reviewed: 19 July 2026

    Who this is for

    Engineering teams shipping containerized applications that need faster delivery without separating security from build, deployment and runtime operations.

    Problems I help solve

    • Vulnerabilities discovered late in delivery
    • Unsigned or untrusted container images
    • Over-privileged Kubernetes workloads
    • Inconsistent infrastructure changes
    • Limited runtime visibility

    How I approach the work

    1. 01

      Map the software path from commit to production

    2. 02

      Set risk-based quality gates

    3. 03

      Sign and attest approved artifacts

    4. 04

      Enforce restricted workload and network policies

    5. 05

      Continuously scan and monitor deployed systems

    Related case studies

    DevOps · 2025

    DevSecOps CI/CD Pipeline with SAST, DAST & GKE Deployment

    I built a fully automated, security-integrated CI/CD pipeline — from commit to production on GKE — with security scanning embedded at every stage, not bolted on at the end.

    Explore

    DevSecOps · 2025

    Kubernetes Security Hardening Playbook

    I built a comprehensive GKE security hardening reference implementation — RBAC design, pod security, network segmentation, admission control, and runtime threat detection, all codified in Terraform and Kubernetes manifests.

    Explore

    DevSecOps · 2024

    Self-Hosted Trivy Scanner on GCP VM

    I deployed a persistent, self-hosted Trivy vulnerability scanner on a dedicated GCP VM — running scheduled scans against all Artifact Registry images and IaC configs, with centralised reporting and CVE trend analysis in BigQuery.

    Explore

    DevOps · 2024

    Full Enterprise CI/CD Pipeline on GCP

    I designed and implemented a production-grade CI/CD pipeline — from commit through automated testing, container builds, infrastructure validation, and multi-environment progressive delivery — entirely on GCP-native tooling.

    Explore

    DevOps · 2024

    Mobile App Publishing Pipeline: Play Store & App Store

    I built and managed the complete end-to-end publishing pipeline for a mobile app — code signing, automated builds with Fastlane, store listing management, crash monitoring, and progressive rollout management for both Android and iOS simultaneously.

    Explore

    Cybersecurity · 2025

    Security Incident Response Automation on GCP

    I built an automated incident response system that contains threats, collects evidence, and notifies the right people — all within minutes of an SCC finding, before a human analyst has even opened their laptop.

    Explore

    DevOps · 2024

    Cloud Load Testing & Performance Baseline System

    I built a repeatable, infrastructure-as-code load testing system on GCP — running distributed k6 load tests from multiple regions simultaneously, capturing performance baselines, and integrating results into CI/CD to catch regressions before production.

    Explore

    DevOps · 2024

    Ephemeral Dev Environment Provisioner

    I built an on-demand ephemeral environment provisioner — developers spin up a full, isolated copy of the production stack (GKE namespace, Cloud Run, Cloud SQL clone) from a Slack command or GitHub PR label, and it auto-tears down after a configurable TTL.

    Explore

    Related writing

    Related services

    Frequently asked questions

    What is DevSecOps in practical terms?

    It means making security checks, artifact trust, infrastructure validation and runtime feedback part of delivery instead of a separate review at the end.

    How do you harden GKE?

    I use least-privilege RBAC, Workload Identity, restricted pod security, default-deny network policy, admission controls, signed images, private networking and runtime detection.

    Which scanning tools appear in your work?

    The portfolio includes Trivy, Snyk, Semgrep, Checkov and OWASP ZAP alongside Artifact Registry, Binary Authorization and Cloud Deploy.

    Can DevSecOps slow delivery?

    Poorly tuned gates can. The goal is to automate repeatable decisions, fail early with useful feedback and reserve manual review for ambiguous or high-risk changes.

    Discuss a technical problem

    Share the system, constraint or control you are working through. I will help clarify the engineering scope and the next useful step.

    Contact Abraham O. Samuel