Amlet — The Global AI Content Registry

The client needed to launch an MVP (Minimum Viable Product) for the AMLET project quickly to validate the idea on the market. The primary requirement was to create a visual identity and logo that were minimal, internationally recognizable, and centered around the letter "a". Additionally, they tasked us with designing and developing the entire web platform (UI/UX, development, and infrastructure), capable of managing content registration and protection in the AI era.

We started by creating a distinctive visual identity, featuring a geometric logo, a functional color palette, legible fonts, and uniform icons. All these elements were chosen to clearly symbolize the link between human creativity and artificial intelligence.

In parallel, we designed a clear and intuitive user interface (UI), intended to facilitate a rapid validation of the MVP. To ensure quality, the interface was developed using wireframes and mockups, focusing on readability, accessibility, and flexibility. The adoption of a dedicated design system also ensures the project's coherence and scalability.

For building the web application, we chose React Router in SSG (Static Site Generation) framework mode, which allowed us to pre-render static pages, ensuring optimal performance. For styling, we opted for vanilla-extract, a CSS-in-TypeScript solution that guarantees complete type-safety and zero runtime overhead. Data validation is managed through Valibot, a modern, lightweight library with full TypeScript support. To build the UI, we integrated react-aria, which adheres to ARIA standards, enabling us to ensure accessibility in our projects. Finally, for client-side state management, we chose Zustand, a minimal yet powerful solution that keeps the code simple and maintainable.

The backend was built with Django Ninja, a modern framework that combines the robustness of Django with an API-first architecture. The choice of Python and Django was driven by the need for integration with the ISCC Framework ecosystem and the vast availability of tools for AI/Machine Learning, content processing, and media fingerprinting. For data persistence, we opted for PostgreSQL 17.5 on Amazon RDS, leveraging the pgvector extension for vector search operations. The infrastructure is containerized with Docker in development, while in production, deployment is on AWS App Runner with infrastructure managed via Terraform. The authentication system is based on API Keys for write operations, with credentials managed through AWS Secrets Manager.

The entire project was delivered on a tight schedule to support an effective MVP launch, which was necessary to validate the project.