No. datascalehr is predictive, not generative. It selects the correct transformation from a finite set of valid mappings and never invents a value. Where a large language model is used, its output only informs a suggestion that a human accepts or rejects. The result is 100% predictable, auditable mapping.
