Extract structured JSON from messy text with Telnyx AI Inference
2 points
1 hour ago
| 1 comment
| github.com
| HN
sona-coffee11
1 hour ago
[-]
I built this as a small Flask example for turning unstructured operational text into structured JSON using Telnyx AI Inference.

The example sends text plus a JSON Schema to a hosted model and returns a JSON object. The default schema is support-ticket oriented, but callers can pass their own schema for leads, incidents, billing requests, etc.

Curious what people think about using schema-guided extraction like this versus more traditional parsers/regexes.

reply