Your question is Synthetic Data for Domain Agents. 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).
How do you structure synthetic data generation and data labeling pipelines to fine-tune specialized domain agents for document processing?