The AI-Native Revolution

A polyglot, headless architecture designed from the ground up to integrate deeply with large language models, RAG pipelines, and modern frontends.

Read the Docs View on GitHub

Core Technical Pillars

Everything you need to build next-generation AI content applications.

Dynamic, Headless Architecture

Content is decoupled from presentation. Build your frontend with any framework while Rosmarium handles the structured data, API delivery, and user roles via comprehensive REST and GraphQL APIs.

Integrated AI Workers

Embedded Python-based intelligence workers run alongside your application. Enjoy native RAG pipelines, automatic tagging, semantic deduplication, and high-performance vector search right out of the box.

PostgreSQL + pgvector

Leverage the world's most advanced open-source relational database. Rosmarium uses pgvector for seamless, scalable similarity searches without the operational overhead of a separate vector database.

High-Speed Content Delivery

Optimized for performance with Edge-compatible delivery endpoints. Advanced caching and robust asset engines ensure your content is served globally with sub-millisecond latency.

Built for Developers.

Fetch content, execute graph traversals, or query vector similarity with simple, intuitive SDKs. Rosmarium stays out of your way so you can focus on building amazing frontends.

View the quickstart guide

fetch-content.ts


import  RosmariumClient  from '@rosmarium/client';

// Initialize the SDK
const client = new RosmariumClient(
  apiKey: process.env.ROSMARIUM_API_KEY,
);

// Perform an AI-powered vector search
const results = await client.content.search(
  query: "How do I configure RAG?",
  mode: 'hybrid',
  limit: 5
);

console.log(results.items);