rust-lang / rust
github.com·1d ago
Developer Tools
Databricks Blog·19h ago·5 min read
Unlock powerful insights and forecasts with Predictive AI Functions
TL;DR
Databricks has introduced Predictive AI Functions in Beta, enabling analysts and business users to perform advanced analytics like forecasting and anomaly detection directly in SQL. This innovation streamlines the analytics process, providing consistent insights without the need for extensive data science resources.
✦ Why It Matters
Start using Predictive AI Functions today to enhance your analytics capabilities without relying on data science teams.
Key Takeaways
- Utilize SQL-native functions for forecasting and anomaly detection.
- Empower business users to perform analytics without data science bottlenecks.
- Integrate predictive functions into existing dashboards for real-time insights.
- Leverage centralized model maintenance for consistent and accurate results.
Full Summary
Data teams often face challenges in answering complex analytics questions, which typically require the expertise of data scientists. This reliance creates bottlenecks, delaying insights for business users.
To address this, Databricks has launched Predictive AI Functions, allowing users to execute advanced analytics tasks such as forecasting, driver analysis, and anomaly detection using SQL-native functions. These functions are powered by state-of-the-art machine learning models maintained centrally by Databricks, ensuring high-quality and consistent results.
Users can invoke these functions directly in SQL, use plain-language queries in Genie, or integrate them into dashboards. The launch includes four new SQL functions and enhancements to the ai_forecast function, improving accuracy in time series projections.
This development empowers business users to derive insights independently, reducing wait times and increasing the number of questions addressed.