Project Charter & Roadmap

Mission Statement

Open Embeddings aims to create a sustainable, standardized way to lower the barrier to entry for new AI agents, scripts, and services while reducing the pressure on content providers from repeated re-embedding operations across multiple models.

Vision

We envision an internet where content discovery is AI-native by default, where:

  • Content providers can expose their material efficiently without bandwidth waste
  • Developers can build semantic search applications without re-embedding costs
  • End users can discover relevant content across platforms and models seamlessly
  • The open internet remains accessible and avoids capture behind walled gardens

Project Goals

When you leave this presentation, you will hopefully understand the thesis and some of you will further:

  1. Define a great Open Format / Spec for content-providers to leverage multiple models
  2. Update commonly used content publishing tools to support the format securely
  3. Generate a corpus of distributed cross-space materials to allow transitions between closed models and open model encoded data
  4. Build the Indexer(s) - Build indexers to allow submissions of open-embeddings from anyone, with degrees of authority built-in

Organizational Structure

  • One person being doing this in their spare time
  • You, if you share the vision