Your question is Implement Named Entity Recognition. 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 working on a text extraction pipeline for a business workflow that needs to identify people, organizations, locations, dates, and product names inside free-form text. The labels are already annotated at the token level, but the text is noisy, with abbreviations, punctuation, and mixed-length sentences.
What is named entity recognition?