We are the research and engineering arm of a group that builds in aerospace and defence, space infrastructure, supply-chain provenance and security. Our work is applied research that becomes intellectual property — held by the Corporation and licensed to the companies that operate on it.
Distributed systems and cryptography are where our published work sits today, because that is the part running in public. They are not the boundary of what we work on.
We build for the group, not for a market. There is no product here to buy and no service to procure. This page exists so that engineers, researchers and regulators can see what we work on and how we work.
Vexidus Corporation operates in aerospace and defence manufacturing, space infrastructure, supply-chain provenance and security. We are the group's research function: our mandate is to work the hard problems underneath those businesses and turn the results into defensible IP.
Consensus under adversarial conditions, and the unglamorous half of it — recovery, catch-up, and what a node does when it has been wrong for six hours without noticing. This is where our published work sits.
Lattice-based signatures in production rather than in a paper: performance, aggregation, and migration paths for systems that cannot stop while they are being changed.
Per-serial traceability, provenance that holds up under audit, and the data problems that arise when a component's history has to be verifiable years after it was made.
Making a record tamper-evident without asking anyone to trust its custodian — chain of custody, document attestation, and verification that survives the issuer.
Not every line of research becomes a product, and not every result is public. What is public is below.
Our distributed-systems work is the part that runs in the open, so it is the part with public artefacts. The protocol specifications are versioned and readable; the three inventions below are the subject of United States patent applications, held by Vexidus Corporation and licensed to the companies that operate on them.
| Invention | Filing | Status |
|---|---|---|
| IntentVM | Non-provisional 19/571,463 | Application filed 18 March 2026 |
| VexBridge | Provisional 63/987,929 | Provisional application filed |
| Atomence | Provisional 63/998,160 | Provisional application filed |
These are applications, not granted patents. No claim of issued protection is made.
Architecture, consensus, token standards, bridge design, governance and identity — written as specifications rather than marketing. docs.vextaris.io
The work is observable rather than asserted: blocks, validators and transactions are public. vexscan.io
Most of what we have learned the hard way is about verification, not about code. These are the working rules, and they are not aspirational — each one was bought with an outage.
A command that succeeds is not a change that landed. We check the thing a user or a peer actually receives. A job that exits zero while silently skipping all of its work looks exactly like a job that ran.
A build output of the right name and plausible size can be unbootable. A backup that arrives on schedule can be empty. The only test of a deployable is deploying it; the only test of a backup is restoring it.
"A broke, and A lacked B" is not a finding. We trace it in the source and cite the line, so that the next engineer can falsify the claim in thirty seconds instead of inheriting it as folklore.
A wrong note is worse than a missing one, because it is believed. When we disprove something we strike it where it lives rather than adding a contradiction somewhere else.
Derivable facts — heights, versions, which machines are healthy — are queried, never recorded. A second copy is a copy that rots, and the rotten copy tends to win because it is the one that gets read first.
We do not keep open vacancies posted when there are none, and there are none today. What follows is honest about the kind of engineer this work suits, so that a speculative note is worth your time or clearly is not.
Consensus, failure modes, and the unglamorous half — recovery, catch-up, and what a node does when it has been wrong for six hours without noticing.
Post-quantum signature schemes in production rather than in a paper: performance, aggregation, and migration paths for systems that cannot stop.
Storage, networking and execution paths where correctness is not negotiable and the profiler is a daily tool.
Aerospace, hardware and controls backgrounds are relevant here. Several of the group's hardest problems are about physical provenance and verification, not about code, and they are usually solved better by someone who has shipped a real part.
People who verify their own work before presenting it, write down why rather than what, and are comfortable saying that a thing they built is wrong.
Speculative introductions are welcome: careers@vexidus.com
Vexidus Labs Ltd is a company registered in England and Wales, company number 13936930, and is the research and engineering subsidiary of Vexidus Corporation, a Delaware corporation.
Registered office: 14/2e Docklands Business Centre, 10–16 Tiller Road, London, England, E14 8PX.
Vexidus Corporation holds the intellectual property, including the patent applications listed above, and licenses it to the operating companies — among them Aerotier in aerospace and defence manufacturing, Cargovate in supply-chain provenance, and Kaarav Technologies in security. Labs performs the research and engineering and generates the IP; it does not hold it.
Vextaris, the Layer 1 network, is a separate identity from the companies that build it. It is intended to be operated by the Vextaris Foundation, a Swiss foundation in formation, which will steward the protocol and issue the native token. Labs does not operate the network.