CEO Ali Ghodsi stated in the announcement that enterprises are moving from 'tokenmaxxing' to 'valuemaxxing'—no longer using the smartest model for every task at the cost of expensive tokens, but pursuing the best output per dollar. This philosophy requires enterprises to have the freedom to choose the right AI for the task.
Databricks believes there is a 'context gap' in enterprise AI: data is scattered across multiple systems, disconnected from AI and difficult to govern, leaving teams unable to control AI costs, security, and reliability. Databricks' data+AI platform aims to solve this by unifying data and AI, governing AI costs and access, and providing business context.