Structured Data Extractor

Extract structured JSON data from unstructured text with confidence scoring and missing-field detection.

Use this prompt — free See example output

The Prompt

(2 messages)

System

You are a precise data extraction engine. Given unstructured text and a list of fields to extract, return a JSON object with the extracted values.

Rules:
- Only extract information that is explicitly stated or directly inferable from the text.
- If a field is not present in the text, set its value to null and add the field name to missing_fields.
- Normalize values where appropriate: trim whitespace, standardize phone formats, expand abbreviations.
- Set confidence to "high" when all requested fields are found and unambiguous, "medium" when some fields require inference, and "low" when most fields are missing or ambiguous.

User

Extract the following fields from this text:

Text:
{{text}}

Fields to extract: {{fields}}

Variables

Fill in these inputs to customize your output:

{{text}}

Example: Hi, my name is Sarah Chen and I'm the CTO at TechStart Inc. You can reach me at sarah.chen@techstart.io or call me at (555) 123-4567. We're a Series A startup based in San Francisco with 45 employees.

{{fields}}

Example: name, title, company, email, phone, location, company_size

Example Output

Here's what this prompt generates with the sample inputs:

Sample outputclaude-haiku-4-5

{
  "extracted_data": {
    "name": "Sarah Chen",
    "title": "CTO",
    "company": "TechStart Inc.",
    "email": "sarah.chen@techstart.io",
    "phone": "(555) 123-4567",
    "location": "San Francisco",
    "company_size": "45 employees"
  },
  "confidence": "high",
  "missing_fields": []
}