VaultTrace
Signed, tamper-evident records of AI sessions: the model the provider declared, the environment and hashes of inputs and outputs, stored in a local append-only ledger. A record-keeping tool, not a compliance certification. Pre-release.
Records you can keep
Forgingverifiable AI
We build AI systems that keep a signed, tamper-evident record of what a session declared and did. Sovereignty and verifiability as engineering, not as a claim of certification.
IndexIndex
What we build, and how far along it is
Two products are public and downloadable today. The rest are in development. We list both, because a company that sells auditability should be auditable itself.
2 published10 in development2 divisions
01
ForgeSynapse
The foundational layer: cryptographic audit, sovereign governance and zero-trust infrastructure.
10 2 published- VaultTracePublished
- ClawLitePublished
- OctoSyncIn development
- V.I.C.T.O.R.In development
- MirrorBreakIn development
- ColdStartIn development
- InboxZeroAIIn development
- PhishNetIn development
- TrustVerifyIn development
- HiveWireResearch
02
Thryvex Labs
Computational biology and operational medicine. Simulation where physical experiments are too slow, too costly or ethically constrained.
2 0 published- TissueForgeIn development
- CasualtyLogicIn development
Available now
Proof you can download
Two of our systems are public today. They are the evidence behind everything else on this page.
ClawLite
A personal AI assistant that runs entirely on your own machine, driven from Telegram. Open source under Apache 2.0: no subscription, no telemetry, no cloud lock-in, and a one-click installer so it is not only for engineers.
Where it matters
Built for environments that cannot afford to be wrong
- 01
Defence
Decision support that stays auditable in disconnected and degraded conditions.
- 02
Health
Computational simulation for trauma research and mass casualty planning, without clinical risk.
- 03
Legal and financial
Signed records that can support an audit or a diligence request. They do not certify conformity.
- 04
High-risk AI
Technical records that can support the logging and documentation work the EU AI Act asks of high-risk systems. They do not meet those obligations by themselves.
Newsroom
What we are building, and why
Thryvex Labs
TissueForge: a million pounds of laboratory, or seventy-two hours of compute
Complex wound research is one of the slowest and most expensive corners of biomedical science. The arithmetic behind that sentence is worse than most people assume.
TissueForge
Thryvex Labs
CasualtyLogic: a hospital rehearses a major incident twenty times a year
In a mass casualty event, the decisions made in the first hours determine most of the preventable deaths. The tools for rehearsing those decisions run at human speed, one scenario at a time.
CasualtyLogic
ForgeSynapse
PhishNet: sharing threat intelligence with people you have no reason to trust
Two organisations both being attacked by the same campaign have every reason to warn each other, and no shared infrastructure to do it through. That gap is where most of the damage happens.
PhishNet
ForgeSynapse
TrustVerify: detecting fakes is the wrong goal
Deepfake fraud took around €860 million in Europe last year, and attacks have risen more than twenty-fold since 2022. Chasing the fakes is a losing race. Certifying what is real is a different problem.
TrustVerify
ForgeSynapse
HiveWire: we are not claiming an AI can make money. We are trying to find out
This is a research programme, not a product, and its own documentation forbids us from selling it until one specific thing is demonstrated. Here is what that thing is.
HiveWire
ForgeSynapse
InboxZeroAI: the one thing you cannot paste into someone else’s cloud
Your inbox is the most sensitive archive you own and the one most people hand over without thinking. InboxZeroAI is an attempt to sort it out without it ever leaving your machine.
InboxZeroAI
ForgeSynapse
OctoSync: one agent is a bottleneck, many agents are a security problem
Running work in parallel is the easy part. Doing it without handing a language model the keys to the host machine is the part most systems skip.
OctoSync
ForgeSynapse
MirrorBreak: what happens when your AI stops disagreeing with you
A model that tells you what you want to hear is not being helpful. It is being useless in a way that feels like help — and the failure gets worse the longer the conversation goes on.
MirrorBreak
ForgeSynapse
ColdStart: the hardest moment with an AI is the first one
Most people leave the majority of a model’s capability unused. Not because the model cannot do it, but because they do not know how to ask. That gap is what ColdStart is being built to close.
ColdStart
ForgeSynapse
V.I.C.T.O.R.: a system that acts, not one that advises
Most assistants answer the question and leave the work with you. V.I.C.T.O.R. is being built to close that gap — and to keep working when the connection does not.
V.I.C.T.O.R.
ForgeSynapse
Verifiable AI proof: the requirement that arrived before most companies were ready
European rules on artificial intelligence are in force. The organisations that lead will not be the ones using the most AI, but the ones able to prove they used it responsibly.
VaultTrace
Contact
Tell us what you need to prove
We do not sell generic products. Every engagement starts with one specific problem, in an environment where trust has to be demonstrable rather than asserted.
contact@forgesynapse.comWrite to us