Agent2Agent (A2A)
Autonomous agents discovering one another and delegating work across hosts over mutually authenticated TLS — including what that means for identity, trust boundaries and auditability once agents start acting on each other’s behalf.
Sovereign AI
I don’t just sell Cyber and AI
I build & test it.
I design, run and maintain a sovereign AI infrastructure on Apple Silicon: a private, multi-node Mac fleet running an Agent 2 Agent architecture accessing local language models, where no data leaves the network.
It is a working environment rather than a demo. That matters, because it has given me hands-on fluency in the standards enterprises are now being asked to evaluate — not a reading knowledge of them.
Standards & architecture
What running it has taught me
Open Knowledge Format (OKF)
Portable, vendor-neutral knowledge and agent memory that is not locked inside a single platform — a live question for any organisation weighing a multi-year commitment to one AI vendor.
Model Context Protocol (MCP)
MCP tooling, local model serving, and retrieval over a governed knowledge vault — how models are actually connected to systems and data, and where the controls need to sit.
Local-first, data-sovereign design
Directly relevant to regulated buyers with data residency, privacy, and security obligations. When inference happens on infrastructure you own, a whole class of exposure never leaves the building.
Why it matters commercially
Two rooms, one conversation.
That combination means I can hold a credible technical conversation with an architect or CISO in the morning, and a commercial one with a CFO or procurement lead in the afternoon, and translate accurately between the two.
Where this is heading
Agentic AI meets regulated buying.
Much of my focus sits at the intersection of agentic AI, data sovereignty, and cybersecurity, particularly what local-first architecture means for regulated organisations across Australia and New Zealand.
It is a fast moving space, and I am always happy to connect and compare notes with others building, evaluating, or buying in it.