TechPrufer
Get started
Databricks
Databricks logoDatabricks · Feature trail

AI Functions in Data Warehouses

2 posts · Aug 2026 – Sep 2026

What this trail covers

This post details the integration and use cases of AI Functions within data warehouses, enabling advanced analytical capabilities and intelligent data processing directly within the warehouse environment. It explores how these functions can be leveraged for tasks such as natural language processing, predictive modeling, and data enrichment, enhancing the utility of the data warehouse for AI-driven insights. The new `ai_decide` function specifically enables fast, low-cost decisions on governed data for model routing, document processing, and agent evaluations.

All Databricks trails
  1. Introducing ai_decide: make fast decisions on your governed dataIntroduces `ai_decide`, a new Databricks AI Function that makes fast, low-cost decisions directly on governed data. It processes unstructured text and returns structured decisions in under a second, suitable for model routing, document processing, and agent evaluations. The post highlights its ability to operate on governed data, implying integration with Unity Catalog for access control and data lineage.
  2. Using AI_Functions in Your Data Warehouse: Top Use CasesThis post details the integration and use cases of AI Functions within data warehouses, enabling advanced analytical capabilities and intelligent data processing directly within the warehouse environment. It explores how these functions can be leveraged for tasks such as natural language processing, predictive modeling, and data enrichment, enhancing the utility of the data warehouse for AI-driven insights.