Your question is Extract Contract Metadata from Text. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You are building an NLP workflow for a legal document system that receives unstructured contracts in plain text and scanned OCR output. The goal is to pull out metadata such as parties, effective date, renewal terms, governing law, and contract type so the records can be indexed and routed downstream. The source text is noisy, with clause headings, boilerplate, abbreviations, and inconsistent formatting across templates.
How would you apply NLP to extract metadata from unstructured contracts?