Spaces:
Running
Running
Welcome to the Tensorus Metadata System
The Tensorus Metadata System is a comprehensive solution for managing metadata associated with tensor data in an Agentic Tensor Database. It enables advanced capabilities such as semantic search, lineage tracking, and version control.
This documentation provides an overview of the system, how to install and configure it, and details about its API and metadata schemas.
Key Features
- Rich Metadata Schemas: Describe tensors with detailed structural, semantic, lineage, computational, quality, relational, and usage metadata.
- Flexible Storage: Supports in-memory storage for quick testing and a PostgreSQL backend for persistent, scalable storage.
- Powerful API: A comprehensive RESTful API for CRUD operations, advanced querying, search, aggregation, versioning, and lineage tracking.
- Data Export/Import: Utilities to export and import metadata in a standard JSON format.
- Security: Basic API key authentication for write operations and audit logging.
- Monitoring: Health check and basic metrics endpoints.
- Enterprise Ready (Conceptual): Designed with enterprise features like JWT authentication in mind.
- Analytics: Example APIs for deriving insights from metadata (e.g., stale tensors, complex tensors, co-occurring tags).
- Dockerized: Easy to deploy using Docker and Docker Compose.
Navigate through the documentation using the sidebar to learn more.