Every agent will claim results. This round funds the place they prove them.
MIZAN built the mechanism for one of the oldest problems in quantitative finance: for forty-six years a track record could not be checked without revealing the strategy that produced it. The mechanism now runs in zero knowledge, in production.
Everything that makes MIZAN worth examining is already public and re-verifiable: the published standard, the research, the live registry of passes and refusals, beginning with our own, and a verifier that runs on your laptop, trusting no one.
116 verdicts stand on the registry, 69 confirmed on Bitcoin; 72 forward days proven, anchored before the outcomes; agent rails live. The standard precedes its first mandate — the way GIPS preceded its first requirement.
Three things you can check right now.
Of 102 credentials on the public registry, 18 pass and 75 fail outright; the other 9 are statistics or not significant. The gate refused Apple’s +272,548% buy-and-hold and an AI net’s +895% on NVDA, and declined to clear its own maker’s flagship: deflated Sharpe 0.68, not significant, published as-is. A referee that will not clear its own founder is one you can expect to hold everyone to the same line.
Systematic capital ($8T) that underdelivers its historical expectation by just 1% loses $80B a year, paid by allocators who had no instrument to check the one number the maths depends on. Every prior fix stopped exactly where that number begins.
Moody’s gave the market a ratings scale and kept the ratings business. Comps: S&P Global, MSCI (~$45B), ICE (~$90B), illustrative of the category, not a valuation. The registry is append-only: the earliest credentials stay earliest for as long as the standard exists.
A referee the market needs cannot be anyone who also trades, which is why we think this seat is worth taking early.
The proof side is live. The capital side is contemplated.
Proofs flow from makers, through the engine, onto an append-only registry the whole market can read. The proof side runs today. The capital side is a contemplated capability: allocation routed against a credential, on licensed partner rails, in a separate entity.
Quants and AI models mint sealed credentials.
Locked gate · zero-knowledge proof · append-only registry · Bitcoin anchor. Era 659bdb67…
Allocators and platforms verify offline in ~81 ms, with the published verifier — exit 0 or refused, no trust in MIZAN required.
Capital routed against a credential, on licensed partner rails. Not live; it scales only as venues come online.
Early here isn’t a discount. It’s a tier you can’t buy later.
Markets settle on one referee.
A market tends to trust one ratings scale, one performance standard: the value is everyone agreeing on the same one, so a second is worth a fraction of the first. Moody’s took its seat in 1909 and has never returned it. Our thesis is that the seat for verified performance opens once, and that it is being taken now.
Append-only means early can never be re-created.
The registry cannot be reordered. The standard’s authorship is fixed. The founding cohort is finite. If diligence questionnaires come to require VTR-1 the way they already require GIPS and an audit, the earliest are already inside, and later capital cannot buy that position back. Late is not a worse entry price. It is a different, permanent tier.
The engine exists. The round buys adoption.
The engine is live (era v12, 659bdb67), the standard is frozen and public, the registry grows, and every claim on this site verifies on your own laptop before we ever speak. The round funds distribution, not existence. Where we are, plainly: 116 verdicts on the registry, 69 confirmed on Bitcoin, 72 forward days proven; agent rails live. The standard precedes its first mandate — the way GIPS preceded its first requirement. If the thesis is wrong, the registry will show it without us.
The registry is append-only. Whether you were early will be a matter of public record, with you or without you.
Request the deck by writing to the founder.
The deck, the one-pager, and the data room are shared directly, not published. Qualified investors may request them by writing to the founder. Conversations are individual, and terms are discussed in them, not here, and not anywhere on this site.
- The live wallEvery credential and every refusal, publicly re-verifiable↗
- The research notesThe receipts, including the failures→
- VTR-1, the standardFrozen, published, free to implement, including by competitors→
- The manifestoWhy this company exists, in six parts→
- The 5-minute demoWatch the gate refuse a beautiful backtest→
Nothing on this page is an offer to sell securities or a solicitation of an offer to buy. Any offering is made only through definitive documents provided directly to qualified investors. MIZAN is a verification and technology provider, not an investment adviser, broker-dealer, or fund. Comparable companies are cited to illustrate the category and are not a valuation. A credential is a seal on a historical or forward record, not a forecast.
Mizan Verification, Inc. · a Delaware corporation · Prove the claim. Enforce the rules. Never reveal the method.