Vaultmark

Inspiration

AI-generated media is becoming a normal part of marketing, design, and content creation, but companies still struggle to prove where an asset came from, how it was generated, and whether it is safe to use commercially. We were inspired by the growing need for AI transparency, provenance, and legal accountability. Vaultmark was created to make AI-generated media not only easy to produce, but also easy to defend during compliance reviews, audits, and intellectual property disputes.

What it does

Vaultmark is a compliance and legal-defense platform for AI-generated media. A user submits a creative brief, generates an image through an AI model, and Vaultmark automatically creates a provenance record for that asset. Every generated file is hashed with SHA-256, linked to a tamper-evident audit chain, and prepared for storage in Backblaze B2 with immutable retention support. The platform also generates two exportable documents: a Compliance Report for regulatory and AI-disclosure requirements, and a Defense Certificate for copyright, likeness, and intellectual-property disputes.

How we built it

We built Vaultmark as a full-stack web application using Next.js (App Router) and TypeScript. The application includes authentication via Supabase, a protected dashboard, an AI generation workflow, asset management, provenance hashing using crypto, and report generation using jspdf. We designed the storage layer around Backblaze B2 using its S3-compatible API (@aws-sdk/client-s3) so that generated assets, manifests, and audit records can be stored durably and immutably. The frontend utilizes Tailwind CSS and shadcn/ui for a modern, responsive design.

Challenges I ran into

The biggest challenge was integrating immutable storage concepts (Object Lock) into a fast-moving hackathon prototype. We also had to adapt our architecture to the deployment environment, which required carefully managing server-side API routes to securely handle Backblaze credentials without exposing them to the client.

Accomplishments that we're proud of

We built a working end-to-end workflow that goes from AI generation to provenance tracking and compliance reporting. The application includes authentication, an asset workspace, sealed compliance status, exportable PDF reports, and a storage architecture designed specifically for Backblaze B2 integration.

What we learned

We learned how important provenance and storage architecture are in AI applications. Building Vaultmark showed us that legal and compliance features can be integrated directly into creative workflows instead of being treated as an afterthought. We also gained experience designing around immutable object storage and audit-friendly metadata.

What's next for Vaultmark

Next, we plan to complete full Backblaze B2 Object Lock integration, add C2PA-compatible provenance manifests, expand support for multiple AI providers through Genblaze, improve similarity and logo detection for brand safety, and add organization-level compliance dashboards.

Built With

• ai
• aws-sdk
• backblaze-b2
• crypto
• jspdf
• next.js
• node.js
• react
• s3-api
• shadcn/ui
• sharp
• supabase
• tailwind-css
• typescript
• vercel