—
GitHub Repositories
// whoami
Cloud & DevOps Engineer working across AWS, Terraform, Docker, and Kubernetes. Background in GIS and remote sensing: I was reading elevation data and satellite imagery long before I read server logs.
7+
Cloud Projects
5+
Years GIS & Infra
1
AWS Certification
// about
My path into DevOps didn't start with a terminal. It started with paper maps. My background is in GIS and remote sensing, where I worked with satellite imagery, elevation models, and multi-criteria decision analysis to answer questions like where's the best place to build, given a dozen competing constraints? I ask a version of that question now about cloud infrastructure too: where a workload should run, how it should scale, what happens when a node fails.
I also modernized legacy spatial data: digitizing paper maps into digital systems, updating road networks, and mapping facilities like bus stops and landmarks. That job taught me something I still rely on. Legacy systems don't get replaced overnight. They get migrated carefully, layer by layer, without breaking what already works.
These days I design and automate cloud-native infrastructure: provisioning with Terraform and Ansible, containerizing with Docker and Kubernetes, building CI/CD pipelines with GitHub Actions and AWS CodeBuild/CodeDeploy. Reliability and observability matter to me because I spent years before this building models that had to hold up against messy, real-world data. That habit didn't go away.
// currently
// skills
// projects
Two ways I've proven this out: an AI service taken from code to a self-healing Kubernetes deployment, and infrastructure-as-code pipelines built on Terraform, plus a spatial-modeling project that shows where the GIS background comes in.
A single application, taken step by step from local prototype to orchestrated production deployment.
A Retrieval-Augmented Generation API: FastAPI accepts a question, ChromaDB retrieves the most relevant
context via semantic search, and a local TinyLlama model (via Ollama) generates a grounded answer. Includes a
dynamic /add endpoint for growing the knowledge base without redeploying.
Packaged the API into a portable Docker image, published to Docker Hub, eliminating environment-specific bugs and making the service deployable anywhere Docker runs.
Automated test pipeline that runs on every push, catching regressions before they reach a deployable artifact, including handling for the non-determinism of LLM output in automated tests.
Deployed the containerized API to a Minikube cluster: Deployments, Services, and label-based routing, with self-healing pod recovery on failure.
Provisioning cloud infrastructure and deployment pipelines as versioned, reviewable code.
Terraform provisions an EC2 instance with S3-backed remote state; GitHub Actions builds a Docker image, pushes it to Amazon ECR, then deploys it to the provisioned instance over SSH: a complete code-to-running-container pipeline.
Terraform provisions a GKE cluster and manages the Kubernetes Deployment/Service directly as code. GitHub Actions authenticates to GCP via Workload Identity Federation, so there are no long-lived service account keys, then builds, pushes, and rolls out the new image automatically.
Role-based Ansible playbooks for automated Linux server setup: Nginx configuration, SSH key management, fail2ban hardening, and static site deployment, agentless, over SSH, across multiple hosts.
Secured a private Maven package repository with AWS CodeArtifact under least-privilege IAM, then automated the build with AWS CodeBuild: a multi-phase buildspec that runs tests before packaging, storing artifacts in S3.
Full deployment lifecycle for a Java web app: cloud dev environment on EC2 via VS Code Remote-SSH, Git/GitHub version control, and AWS CodeDeploy lifecycle hooks (BeforeInstall, AfterInstall, ApplicationStart) for zero-touch releases.
Deployed a production bookkeeping platform on AWS ECS Fargate, applying containerised, serverless infrastructure principles to a live application, not a tutorial exercise.
Where the GIS background shows up directly.
A multi-criteria spatial suitability model for identifying optimal dam construction sites across Ekiti State, Nigeria. Integrated geologic mapping, land cover classification, SRTM digital elevation data, and hydrological/drainage data into a single decision framework, using the Analytic Hierarchy Process (AHP) to weight competing criteria. It's a similar kind of weighted-tradeoff reasoning to what goes into choosing cloud regions, instance types, or architecture patterns under competing constraints like cost, latency, and resilience.
// ai evaluation & annotation
FOUNDATIONAL — BUILDING NOWIn July 2026 I started deliberately building a freelance practice in AI data annotation and output evaluation: rating and comparing AI-generated outputs against structured criteria across platforms including CrowdGen, Mindrift, Mercor, DataAnnotation, and AfterQuery. I'm early in this. Onboarding, qualification exams, and early tasks, not a long track record yet. Still, it isn't a random pivot. Below is the pattern that got me here.
01
to site a dam. My GIS thesis, using AHP.
02
to choose infrastructure. My DevOps work.
03
to judge AI-generated outputs. This niche.
Across 90+ image-comparison tasks, a working framework emerged for judging AI outputs against a prompt. It's prioritized in this order, ahead of raw aesthetics:
Onboarding and building a track record across several evaluation platforms in parallel, treating each qualification exam as a chance to sharpen the same methodology rather than a one-off gate to clear.
I'm treating this as a foundational layer, not a finished credential. I'll update this section as the track record builds. What's already true is that the analytical habit behind it isn't new. It's the same structured, criteria-first thinking that shows up in the GIS work and the infrastructure work on this page too.
// github activity
LIVE—
Public Repos
—
Total Stars
—
Last Push
// experience
July 2026 — Present
Evaluating AI-generated outputs across CrowdGen, Mindrift, Mercor, DataAnnotation, and AfterQuery, alongside continued self-directed cloud/DevOps project work (Terraform, Kubernetes, CI/CD).
Oct 2021 — Sep 2023
Aug 2022 — Aug 2023
Society role alongside full-time employment: public outreach on GIS careers and contributions to an urbanisation and road-mapping project covering Akure Metropolis.
2024 — 2025
Stakeholder-facing and operational roles held during the DevOps upskilling period. Full detail on request.
// certifications
Amazon Web Services · 2024
CompTIA
AZ-900
// contact
Open to Cloud, DevOps, and Infrastructure roles. Also always happy to talk about where spatial data thinking meets systems design.
