A semantic file search system that uses vector embeddings to enable natural language queries across your documents.
SFS allows you to index files and search their content using natural language queries. Instead of searching for exact keywords, you can ask questions like "what does this say about deployment?" and get relevant results.
| Component | Description | Link |
|---|---|---|
| sfs-api | FastAPI backend with vector search and file management | GitHub |
| sfs-desktop | Cross-platform desktop GUI built with Tauri + React | GitHub |
| sfs-cli | Command-line interface to use SFS | GitHub |
Note: Desktop and CLI don't provide the same features
Lightweight and easy to use, the CLI provides a command-line interface for uploading and searching documents. Has more features such as Watch to auto upload files when they are modified.
More intuitive than the CLI, the desktop app provides a graphical interface for uploading and searching documents.
See the sfs-api README for deployment instructions using Docker Compose.
- Semantic Search: Query documents using natural language
- File Storage: S3-compatible storage server
- Async processing: Distribute work for parallel processing
- Encryption: Server-side encryption for stored files
- Backend: Python, FastAPI, Minio, Qdrant
- Frontend: React, Tauri, TypeScript
- CLI: Go, Cobra
See individual repositories for license information.