Neil Smith

The work

I’m a senior cloud security engineer currently working at ONYX Insight — securing critical infrastructure that monitors and maintains wind turbines across tens of thousands of edge-connected devices. The environment is OT-adjacent: SCADA-integrated, safety-critical, and running AI/ML systems for predictive maintenance at scale.

Before that: a regulated financial services environment (Nottingham Building Society), an incident that involved law enforcement and a forensic response team (Barbon Insurance Group), and fifteen years building infrastructure and security capability at a university from the ground up.

The career is coherent even if it doesn’t look it at a glance: every role has been the same problem at a different scale. How do you assure that a complex system with many inputs, many people, and many failure modes is actually doing what it claims? And how do you prove it?

Why AI security

The industry is moving fast and the governance is thin. Organisations are deploying AI into production environments without meaningful threat models, without tested controls, and without the frameworks to know when something has gone wrong.

That gap is where the serious work is. Not AI security theatre — actual adversarial testing, actual governance frameworks applied to running systems, actual incident response planning for when a model behaves in ways it shouldn’t.

My specific focus is the intersection of AI systems and critical infrastructure. AI deployed into OT environments, predictive maintenance pipelines, industrial telemetry — where the threat landscape is least understood and the consequences of failure are most serious.

Credentials

• GIAC GAIPS (AI Platform Security) — in progress (2026)
• GIAC GICSP (Industrial Cyber Security Professional) — planned
• GitLab Associate
• Eligible for SC Clearance

The platform

OMEGA Core is the research platform behind this work — a dedicated local AI and security engineering environment running Ollama, FastAPI, PostgreSQL, Qdrant, InfluxDB and Grafana on a MINISFORUM AI X1 Pro-470. Everything published here is grounded in real systems, tested controls, and documented findings from that platform.

Working together

If you’re looking for AI security architecture review, governance framework design, or specialist input on AI systems in critical or regulated environments — get in touch.