Data Extraction Pipeline
hardMulti-Agent8 tests
Your data team gets unstructured text dumps from partners and manually copies info into spreadsheets. Build an ETL pipeline: one agent extracts the key fields, another normalizes them into clean JSON. No markdown code blocks, just raw JSON.
Agent Prompts
Extractor
Extracts key information from unstructured text.
Normalizer
Normalizes the extracted data into a clean JSON structure.