Founder, ObraVera | Full-Stack Developer | Author
I build content provenance infrastructure for the generative AI era — fingerprinting and verification tools that let publishers, journalists, and creators establish authorship and identify unauthorised use of their work.
A registry-based platform for verifying human authorship and identifying creative works that have been scraped, ingested, or reproduced by AI systems. ObraVera's defining capability is retrospective detection — working from content fingerprints rather than embedded metadata or pre-applied watermarks, so works already in circulation remain protectable.
Text
- ISCC (International Standard Content Code) and Winnowing-based fingerprinting
- Content-derived identifiers, generated from the work itself — the source content
- is not modified — registration produces a fingerprint from the work itself.
- Back-catalogue registration: archives that pre-date the AI era can still be fingerprinted and registered
- Aimed at journalism, books, and long-form text where the asset already exists in the wild
Audio
- Acoustic watermarking via audiowmark (C++) and Meta's AudioSeal
- End-to-end tested through analogue gaps (speaker → microphone → upload), with confidence scoring
- Live verification portal (currently restricted to beta testers): verify.obravera.com
Stack: Django · Docker · Python · HTMX · C++ library integration · deep learning model deployment
Forward-looking schemes (C2PA, watermarking, "Proven Human" badges) protect what gets published tomorrow. ObraVera identifies what was taken yesterday — and registers what exists today.
DAW plugin in development. Session-level attribution and royalty workflows for music producers, engineers, and session musicians.
[Python] [Django] [JavaScript] [HTMX] [Docker] [Bash] [Git] [C++ integration]
- Author: Elements of Risk and The Horn Mansion — bradyridgway.com
- obravera.com — Company site
- verify.obravera.com — Live verification portal
- ObraVera on Substack — Writing on AI, authorship, and provenance
- bradyridgway.com — Author site
"Start small, test often, and iterate."