AivexaNewsSearch
AI news for builders and product teamsChecked every hour

Unlock powerful insights and forecasts with Predictive AI Functions

Collected Oct 15, 2026

Databricks announced a new capability called Predictive AI Functions, aimed at data teams that are regularly asked forward-looking business questions such as what sales will be next quarter. The company framed the launch around the idea that these questions come up repeatedly and that answering them should not require leaving the data platform.

The core pitch is that forecasting and insight generation become functions available inside the workflows data teams already use, rather than a separate modeling project. Databricks describes the result as powerful insights and forecasts, positioning the feature as a way to unlock predictive output from existing data work.

This suggests the functions are meant to lower the barrier to producing forecasts: instead of standing up and maintaining bespoke models, a team can invoke a predictive function the same way it would any other transformation in its pipeline. That is a meaningful convenience for analysts and analytics engineers who are comfortable in SQL or notebooks but not necessarily in machine learning engineering.

In practice, the value of such an approach depends on how the functions handle seasonality, missing data and changing trends, and on whether their outputs are explainable enough for business stakeholders to trust. A likely trade-off is that a general-purpose predictive function will be easier to adopt but less tunable than a purpose-built model when the forecasting problem is unusual or high-stakes.

Why it matters: data teams are frequently the first stop for forecasting requests, and any feature that shortens the path from question to answer changes how that work is staffed and scheduled. For developers and product teams, the practical question is whether Predictive AI Functions can replace some ad hoc modeling work or will sit alongside it as a fast first pass.

Read at Databricks

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

Data teams all get asked the same questions: "what will sales be next quarter?,"...