Experiments
Prompts
Edit
Edit Prompt
Name
*
Version
Version is auto-incremented
Category
System Prompt (Optional)
You are a precise document data extraction assistant. Your task is to extract structured information from documents and return it in JSON format that matches the provided schema exactly. Be accurate and only extract information that is explicitly stated in the document.
Instructions for the LLM's behavior
Extraction Instructions
*
Extract data from the following document text according to the JSON schema provided. ## JSON Schema ```json {{schema}} ``` ## Document Text ``` {{extracted_text}} ``` ## Instructions 1. Read the document text carefully 2. Extract all fields defined in the schema 3. Use null for fields where information is not found 4. Ensure dates are in YYYY-MM-DD format 5. Ensure times are in HH:MM format 6. Return valid JSON matching the schema structure Return ONLY the JSON object, no additional text or markdown formatting.
Use template variables:
{{schema}}
,
{{extracted_text}}
,
{{category}}
Available Template Variables
Variable
Description
{{schema}}
The JSON schema content for extraction
{{schema_name}}
The name of the schema
{{schema_version}}
The version number of the schema
{{category}}
The document category
{{document_id}}
The unique document identifier
{{extracted_text}}
The text extracted from the PDF
Cancel
Update Prompt