Security-first AI from operators who’ve run the infrastructure, the business, and the finance — local models doing frontier-grade work, your data never leaving the building.
AI is moving fast. We’re already where it’s going.
Founded & led by Ian MacDonald, CTO · 100% of client data stays in your environment
Research, in the open
We publish what we’re building, breaking, and learning — the evidence behind every claim on this page.
From Zero to Tokens: ROCm 7.0.2 Quickstart on Cloudrift’s 8-GPU Node
Read article →How to Force-Enable Claude Opus 4.6 [1M] Context with a Max Subscription
Agents & ToolingHow We Fixed a Broken Claude Code Session (and Built a Tool for Everyone)
Self-Hosted AIStrix Halo on Ubuntu looks great
Knowledge SystemsThe Knowledge Flywheel: How AI-Powered Wikis Forge Smarter Teams and Build the Future of Work
The Long GameHidden Risks of Valuable Human Capital
Build NotesCloudflare tenant bypass in the wild
Agents & ToolingA Story About ChatGPT GPT-5
Agents & ToolingPrompt Shaping
The Long GameAccelerating Innovation: Balancing Risks and Rewards in AI and Deep-Sea Exploration
Self-Hosted AIDebian: The Secret Weapon to a Faster and More Efficient Agent Development
The Long GameBenjamin Franklin
The Long GameClarke’s three laws
What we do — and the proof
Data Privacy
Your data never leaves your environment. Local-model guardrails do frontier-grade work on your hardware. Privacy isn't a setting; it's the architecture.
Read the research →AI Tools
Agents that ship work, not demos — harnessed around real workflows, measured on output.
Read the research →Identity
Know exactly who's in your systems — identity-first access for people and agents.
Read the research →Network Security
Security that assumes the breach — segmented, monitored, least-privilege.
Read the research →Business Continuity
Tested recovery, not hopeful backups — snapshot / restore / failover proven against real outages.
Read the research →About Ian
Ian MacDonald has spent a career where most AI conversations stop short: at the intersection of infrastructure, business operations, and finance. Models are easy now; making them safe, private, and genuinely useful inside a real business is the hard part — and it’s the part Ian has been doing hands-on for years. He leads Netstatz as founder and CTO, building AI that ships fast, respects data, and earns its place in the workflows that actually run a company.
Currently open to senior AI-leadership roles for organizations serious about vertical, privacy-first AI integration — let’s talk before the calendar fills.
Three ways to work with us
$500/month buys you an AI practice, not a chatbot.
Two meetings a month, real direction, and project work when it fits what we're building. Privacy-first, vendor-honest, shipping every cycle.
Get started — $500/moThe weekly call where AI actually gets discussed.
Open, candid, technical. Bring a problem, leave with three approaches. The community we wished existed — so we built it.
Join the weekly callTalk to Ian — before he gets too busy.
Hiring an AI leader who knows infrastructure, finance, and how to ship? Ian is open to the right senior role. The window won't stay open long.
Talk to Ian