eidos: an extensible local-first personal data management framework with AI-powered data interaction

What it solves

Eidos provides a framework for personal data management, allowing users to organize and store information using a combination of Notion-like documents and databases. It addresses the need for local-first, private data storage that remains extensible and AI-enhanced.

How it works

The system transforms SQLite into a personal pocket database. It runs entirely on the local machine for offline support and performance. It features an extension system where users can write custom code in TypeScript, JavaScript, or Python to create UI components (Blocks) or data processing logic (Scripts), which can also be used as tools by integrated LLMs.

Who it’s for

Users who want a private, local-first alternative to Notion for managing personal data, as well as developers who want to build custom tools and data processing workflows directly into their personal knowledge base.

Highlights

• Local-first: All data is stored in SQLite and runs offline for speed and privacy.
• AI Integration: Built-in LLM capabilities for translating, summarizing, and interacting with personal data.
• Extensible Architecture: Supports custom UI blocks and scripts that can be manually written or AI-generated.
• Open Format: Data is stored in an open SQLite format, ensuring users maintain full ownership of their raw data.