Connector Improvement: Feature Request: Structured Destination Schema Metadata for AI Consumption
Not plannedAs a Fivetran user leveraging the GitLab connector and evaluating PagerDuty, we request access to destination schema metadata in a structured, machine-readable format (e.g., JSON Schema or a dedicated API endpoint).
Currently, while ERDs are helpful for manual review, they are not suitable for automated processing or AI tools. Specifically, we need:
- Structured schema definitions and descriptions for destination tables (like those created by the GitLab and PagerDuty connectors).
- Explicit definition of primary and foreign key constraints within the metadata.
- Clear indication of data relationships between tables (similar to the Google Workspace schema example https://support.google.com/a/answer/7230050?hl=en&src=supportwidget0&authuser=0).
Providing this structured metadata would significantly enhance our ability to programmatically understand and integrate Fivetran data using AI-powered tools.
-
Official comment
Hi Jason,
Thank you for reaching out and sharing your thoughtful request with us.
At this time, the feature you’ve described is not on our roadmap. While it’s not currently planned, we are always open to reconsideration as we continuously evaluate our roadmap based on user feedback and emerging needs. We encourage you to share this request within our community and have others upvote it, as community interest plays a crucial role in shaping our priorities.
We do want to ensure that you have a path forward. We currently share the metadata tables in the destination which can be transformed downstream into flattened JSON structures. Please do reach out if you need any help at all.
Thank you again for your input, and do feel free to reach out if you have any more ideas or questions.
Best,
Meera
Please sign in to leave a comment.
Comments
1 comment