Gideon Warui

Cloud Platform & DevSecOps Engineer · AI Infrastructure Builder · Nairobi, Kenya

I build production-grade platforms and infrastructure across cloud, DevSecOps, data, and AI systems. In enterprise roles, that has meant owning Azure infrastructure, AKS clusters, CI/CD pipelines, observability, incident response, and cost optimization. In product and data roles, it has meant building data pipelines, analytics systems, and AI-powered workflows that teams actually use.

My background is intentionally layered: freelance data science and BI projects after university, then building a startup data function from scratch, then consulting across NGO, fintech, and enterprise environments, and now operating at the intersection of cloud platform reliability and AI infrastructure. That progression is why this site covers both platform operations and AI systems, not one or the other.

The primary goal of this portfolio in 2026 is to document a public AKS AI Infrastructure Lab: a 45-week curriculum (~80 labs) paired with a two-posts-per-week writing cadence (~84 posts) covering Kubernetes internals, GPU serving, vLLM/Triton, RAG systems, autoscaling, chaos testing, and FinOps. Each lab is designed to be built in 1–2 days, isolated by namespace, instrumented, measurable, and teardown-friendly.

This site is the evidence layer for that work: real configurations, screenshots, measurements, and production-style tradeoffs. I care less about broad tutorial coverage and more about proving claims with data, code, and reproducible experiments.

Quick facts
Location
Nairobi, Kenya · Remote
Focus
Cloud Platform · DevSecOps · AI Infra · FinOps
Current build
2026 AI Infra Lab · 45 weeks · ~80 labs
Publishing cadence
2 posts/week · ~84 code-heavy posts
Lab stack
AKS · vLLM · Triton · KEDA · Qdrant · pgvector · Prometheus · Grafana · Loki
Status
Available for opportunities

How I work

01

Instrument first

Never optimize what you haven't measured. Every lab starts with instrumentation and a baseline, then moves into experiments. Numbers before claims.

02

Everything committed

Helm values, dashboards, load fixtures, and notes. If it is worth claiming, it should be reproducible from the repository.

03

One concept per lab

Narrow scope wins. Each lab proves 1–3 concepts, captures real measurements, and ends with a clear production tradeoff and a teardown path.

Education

Bachelor of Science in Mechatronic Engineering

May 2017 – May 2022

Dedan Kimathi University of Technology · Specialization: Automotive Mechatronics

Mechatronics sits at the intersection of mechanical engineering, electronics, control systems, and computer science — the discipline of building intelligent machines. The automotive specialization covered ECUs, embedded control systems, CAN bus architecture, ADAS fundamentals, and vehicle diagnostics. Five years of learning to think in systems: how hardware, software, and physics interact under real-world constraints.

Final-year project: An augmented reality system for vehicle technicians — overlaying maintenance procedures and component identification onto live camera feeds, reducing service errors and accelerating apprentice onboarding on the workshop floor.

Certificate in Data Science

Feb 2022 – Nov 2022

Moringa School · Part-time (9 months)

Python for data analysis, statistical modeling, and machine learning. Covered supervised and unsupervised learning, deep learning fundamentals, neural networks, and large-scale data processing with PySpark. I took it part-time during the final semester of engineering — a deliberate pivot from physical systems to data systems.

Certificate in DevOps Engineering

Aug 2023 – Dec 2023

Moringa School · Intensive (5 months)

Linux systems administration, containerization with Docker, Kubernetes orchestration, CI/CD pipeline design with GitHub Actions, infrastructure as code with Terraform, and cloud operations on AWS. The technical foundation for everything that followed — learned the full deployment lifecycle from code commit to production cluster.

Cloud Computing Program

Jan 2024 – Jun 2024

ALX Africa · Cloud Practitioner Track (6 months)

Cloud architecture principles, AWS services across compute, storage, networking, and security, serverless patterns, IAM, and cloud-native design. ALX Africa's intensive practitioner program — structured around real-world scenario labs — strengthened the cloud architecture layer of my transition from data systems into platform engineering.

Let's talk platform & AI infrastructure

Open to platform engineering, DevSecOps / SRE, AI infrastructure, and architecture-focused consulting.