By Micky Irons, founder of Mickai.
The problem the public sector cannot design away
Mickai is a sovereign intelligence operating system that regulated businesses own and run inside their own walls. Government departments, local authorities, healthcare trusts and defence bodies hold the most sensitive data in the country. They also face the strictest rules on where that data can travel. The result is a hard limit that most AI vendors quietly ignore: a great deal of public sector data legally cannot be sent to a public cloud model, no matter how capable that model is.
We built for that limit rather than around it. Our system runs entirely on the customer's own hardware, on premises and air gapped, with zero data egress and no public cloud round trip. Nothing leaves the building. That is not a configuration option we bolt on for nervous buyers. It is the design premise of the whole platform, and it is built and live today, not a concept or a roadmap.
The regulatory pressure is not abstract. UK GDPR special category data, the NHS Data Security and Protection Toolkit, the EU AI Act high-risk classification, ITAR and EAR, the NIS Regulations and the US CLOUD Act each constrain where public sector and regulated data can be processed. The CLOUD Act alone is enough to make many bodies uncomfortable with any architecture where data crosses a jurisdictional boundary or sits under foreign legal reach. We remove that question entirely, because the data never moves.
What we run, and why it is reproducible
We run about fifty specialist models, twenty five domain and twenty five operational, with cross-model routing under a deterministic arbiter. That last point matters more to a public sector buyer than raw model size. Because routing is deterministic, outputs are reproducible. The same inputs produce the same result, which is precisely what an auditor, a tribunal or a public inquiry expects and what a probabilistic black box cannot offer.
We do not name the sovereign models beyond that description, and we do not need to. The capability is delivered as studios, each with a Greek name and a serious function. Nemesis handles fraud and anti money laundering. Plutus covers finance and FP&A. Tyche does underwriting. Prometheus runs forecasting. Iris handles customer service. Nomos covers compliance, Astraea covers legal, Panacea covers clinical work, Pythia is business intelligence and Aletheia is audit. Vinis handles voice. The Agentic Marketing Team, Trust Agent as the perimeter, and OAR-as-a-Service complete the set. A public body can adopt the studios relevant to its mandate and leave the rest.
The Open Audit Record
Public sector accountability lives or dies on the audit trail. Our answer is the Open Audit Record. Every consequential action is signed under post-quantum cryptography, using FIPS 204 ML-DSA-65 with ML-KEM-768, and hash-chained into a tamper-evident, append-only ledger.
The important property is independence. Anyone can verify that ledger offline, for decades, without trusting us. A future auditor does not need our servers, our goodwill or our continued existence to confirm what happened and when. For a body that must answer to Parliament, to a regulator or to the public, that is the difference between a claim and a proof. We chose post-quantum signatures because records the state keeps must outlast the cryptography of the moment they were created.
Attestation across sites without a central server
Many public sector estates are distributed. Multiple sites, multiple trusts, multiple fielded units, often with no reliable shared connectivity. Pantheon, our post-quantum Layer 1 currently on testnet, gives multi-node attestation across those fielded units with no central server. Each unit can attest to the others without a single point of trust or a single point of failure, which suits an estate that must keep working when a link goes down and must never depend on one privileged node.
The intellectual property behind the platform
The architecture is protected. We hold 104 filed UK patent applications, roughly 2,340 claims, across 13 invention families, owned by Mickai LTD, with named inventor Mickarle Sean Junior Wagstaff-Irons. These are filed, not granted. We are precise about that, because filing is what establishes priority and builds a prior-art moat, and precision is what a public sector procurement team is entitled to expect.
Mickai LTD is a UK company, Companies House number 17166618, with Birmingham manufacturing secured. That last point matters to any body with a domestic supply chain preference or a sovereignty mandate of its own. The hardware and the platform are British, and the manufacturing base is in place.
The market we serve, and the one reading over our shoulder
The sovereign AI market is roughly USD 40 billion in 2025, rising to about USD 148 billion by 2032. Around 0.85 million UK businesses, about 15 percent, and roughly 5 million across the EU legally cannot send data to public cloud AI. Public sector bodies sit squarely inside that population, alongside banks, insurers, hospitals and defence suppliers. The drivers are consistent: PRA model-risk expectations under SS1/23, UK GDPR special category data, the NHS Data Security and Protection Toolkit, the EU AI Act high-risk classification, ITAR and EAR, the NIS Regulations and the US CLOUD Act.
Our strategy has two sides. We sell sovereign AI to the regulated firms and public bodies the public cloud cannot lawfully reach. We also license the patented stack to the platforms that want to reach them. Our internal analysis maps 196 companies and 311 patent-company pairs as potential licensees, including Microsoft, AWS, NVIDIA, Google, Adobe and IBM. That is potential-licensee sizing, not a signed book and not an infringement claim. We are an ally to the AI majors, not a rival to them. A platform that adds a sovereign layer instantly reaches a regulated market it cannot serve today, and public sector procurement is a large part of that market.
What this means for a public sector buyer
For a government or public sector team, the calculus is straightforward. The AI capability is real and running now. The data never leaves your estate, so the residency and CLOUD Act questions resolve before they are asked. The outputs are reproducible, because routing is deterministic. The audit trail is independently verifiable for decades under post-quantum signatures. The supply chain is British and the manufacturing is secured. Those are the properties a public body needs to defend a decision, not the properties a vendor invents to close one.
Our pre-seed round is opening soon, and we welcome inquiries from interested partners by email at micky@mickai.co.uk or on LinkedIn.
FAQ
Does any public sector data leave our premises when we use Mickai?
No. The platform runs entirely on your own hardware, on premises and air gapped, with zero data egress and no public cloud round trip. Nothing is sent to an external model or service, which is why residency, UK GDPR special category, and US CLOUD Act concerns do not arise in the first place.
How can we audit AI decisions after the fact?
Every consequential action is signed under post-quantum cryptography (FIPS 204 ML-DSA-65 with ML-KEM-768) and hash-chained into a tamper-evident, append-only ledger, the Open Audit Record. Anyone can verify it offline, for decades, without trusting us, which is what makes it suitable for regulatory, tribunal and public inquiry scrutiny.
Is this a finished product or a plan?
It is built and live today. We run about fifty specialist models under a deterministic arbiter, organised into studios covering fraud and AML, compliance, legal, clinical, audit and more. Pantheon, our multi-node attestation layer, is currently on testnet. The company is Mickai LTD, Companies House 17166618, with Birmingham manufacturing secured.
Written by Micky Irons, founder of Mickai. Originally published at https://mickai.co.uk/articles/government-public-sector. More from Mickai at mickai.co.uk.
Top comments (0)