mirror of
https://github.com/run-llama/template-workflow-data-extraction.git
synced 2026-07-01 21:34:19 -04:00
Adrian Lyjak
46 lines
1.2 KiB
JSON
46 lines
1.2 KiB
JSON
{
|
|
"extract": {
|
|
"product_type": "extract_v2",
|
|
"configuration_id": null,
|
|
"data_schema": {
|
|
"type": "object",
|
|
"properties": {
|
|
"document_type": {
|
|
"type": "string",
|
|
"description": "An overarching category for the type of document (e.g. invoice, purchase order, etc.)"
|
|
},
|
|
"summary": {
|
|
"type": "string",
|
|
"description": "A 2-3 sentence summary describing the content of the document"
|
|
},
|
|
"key_points": {
|
|
"type": "array",
|
|
"items": { "type": "string" },
|
|
"description": "A list of key points or insights from the document"
|
|
}
|
|
},
|
|
"required": ["document_type", "summary", "key_points"]
|
|
},
|
|
"tier": "agentic",
|
|
"extraction_target": "per_doc",
|
|
"cite_sources": true,
|
|
"confidence_scores": true
|
|
},
|
|
"classify": {
|
|
"product_type": "classify_v2",
|
|
"configuration_id": null,
|
|
"rules": []
|
|
},
|
|
"parse": {
|
|
"product_type": "parse_v2",
|
|
"configuration_id": null,
|
|
"tier": "agentic",
|
|
"version": "latest"
|
|
},
|
|
"split": {
|
|
"product_type": "split_v1",
|
|
"configuration_id": null,
|
|
"categories": []
|
|
}
|
|
}
|