I’m a warehouse worker in Europe who codes at night.
Last week I shipped Crovia Trust: a fully offline, open-source (Apache 2.0) engine that turns any provenance log (FAISS, MIT DPI, Hugging Face, etc.) into verifiable attribution + fair royalty bundles. No blockchain, no tokens, no SaaS or cloud dependency.Key points Works 100 % offline – Merkle proofs generated locally in <30 seconds
Tested on the public MIT Data Provenance dataset: 3 718 receipts → €1 M simulated payout with Gini 0.23 (very equitable)
Produces compact CEP.v1 evidence bundles ready for EU AI Act Annex IV / model cards 2026
Core is frozen, fully auditable, zero external deps beyond standard Python
Demo tarball + verification script downloadable, runs on any laptop
Repo & 30-second demo: https://github.com/croviatrust/crovia-core I built it because I’m tired of seeing creators and small data providers get zero while labs raise billions on their backs. Happy to answer technical questions, help integrate it, or just hear brutal feedback.
crovia•2mo ago
Last week I shipped Crovia Trust: a fully offline, open-source (Apache 2.0) engine that turns any provenance log (FAISS, MIT DPI, Hugging Face, etc.) into verifiable attribution + fair royalty bundles. No blockchain, no tokens, no SaaS or cloud dependency.Key points Works 100 % offline – Merkle proofs generated locally in <30 seconds
Tested on the public MIT Data Provenance dataset: 3 718 receipts → €1 M simulated payout with Gini 0.23 (very equitable)
Produces compact CEP.v1 evidence bundles ready for EU AI Act Annex IV / model cards 2026
Core is frozen, fully auditable, zero external deps beyond standard Python
Demo tarball + verification script downloadable, runs on any laptop
Repo & 30-second demo: https://github.com/croviatrust/crovia-core I built it because I’m tired of seeing creators and small data providers get zero while labs raise billions on their backs. Happy to answer technical questions, help integrate it, or just hear brutal feedback.
Thanks for looking.