Anaplan CEO Describes AI Changes to Enterprise Software Architecture

The executive described a three-layer structure. The top layer consists of large language models that process natural language requests. A middle layer performs exact, auditable computations required for enterprise decisions. The bottom layer converts those results into commands for other business systems.

Vendors whose main offering is data visualization or simple query interfaces face reduced demand, according to the executive. Tools that previously made data easier to view or navigate are now competing with language models that can summarize the same information.

Vendors that maintain systems producing identical results from identical inputs may see increased value. The executive listed planning engines, human capital management systems, and customer relationship management systems as examples of products that perform such calculations.

The executive noted that companies with unique computational models on these systems gain an additional advantage. Replicating those models in competing platforms requires more than data transfer and creates a barrier for customers considering alternatives.

The executive added that investment should shift from interface design toward deeper computational capabilities and stronger governance controls.