AI-ready data
For data to be AI-ready, it generally needs to be described in a machine-readable format.
This should address:
- the way by which the data is accessed (API)
- the shape of the data (Objects, Tabular, Files, Images…)
- the context in which the data can be used (license, provenance, usage permissions and restrictions)
Machine Readable Data
Standards
| Specification | Application |
|---|---|
| OpenAPI Specification | API description |
| DPROD | Data Products and data services description |
| Croissant 1.1 Format | ML dataset metadata standard |
Corresponding Tools
| Tool | Application |
|---|---|
| OpenAPI MCP Server | OpenAPI MCP |
| FastMCP | OpenAPI MCP |
| Éclair | Dataset MCP |