NLU System Design — Intent & Entity Architecture Report

I will assess your NLU requirements and deliver a written system design report covering your intent taxonomy, entity schema, training data strategy, NLU platform selection (Rasa, Dialogflow, custom transformer, or BERT-based), and a realistic implementation roadmap. NLU system design mistakes made before any development begins — an overly granular intent taxonomy that confuses the model, an entity schema that doesn't capture the distinctions your downstream logic needs, or a platform choice that can't handle your language or domain — are far more expensive to fix after training data has been collected and a model has been trained than before.

The report covers your use case and conversation types, a recommended intent taxonomy, entity schema design, training data volume estimate per intent, platform comparison for your requirements, and a phased development roadmap.