GitHub user richardchen874-sys added a comment to the discussion: [Ideas]  
AI-Powered Parser Function for Unstructured Data

An AI-powered parser function for unstructured data will probably spend more in 
repair and retry loops than in the first successful extraction.

I would expose parser runs as structured telemetry: input size, detected 
format, model used, schema validation result, retry count, and cost per 
accepted row/document. That gives users a way to compare models without turning 
the SQL function into a black box.

I am testing an OpenAI-compatible multi-model API layer around official Chinese 
models, with a lot of attention on usage/cost visibility and fallback. Would 
Cloudberry users need parser-level cost estimates before execution, or only 
post-run accounting?

GitHub link: 
https://github.com/apache/cloudberry/discussions/1442#discussioncomment-17903568

----
This is an automatically sent email for [email protected].
To unsubscribe, please send an email to: [email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to