Starburst offers AI SQL functions that allow data analysts to apply AI-powered text analysis — including sentiment analysis, classification, grammar correction, translation, and data masking — directly within standard SQL queries. These functions are available in both Starburst Galaxy and the Starburst Enterprise Platform.

Called Starburst SQL Functions (or Starburst AI Functions), these capabilities embed generative AI into familiar SQL syntax so that analysts can analyze and transform unstructured text without moving data to a separate tool or writing specialized AI code. For example, a function like `starburst.ai.analyze_sentiment()` takes a text string and an AI model alias as inputs and returns a sentiment label (positive, negative, neutral, or mixed). A `starburst.ai.classify()` function categorizes input text according to user-supplied labels. Multiple LLMs are supported as the underlying model, including AWS Bedrock and OpenAI-compatible services. These functions were introduced as part of Starburst's AI Workflows suite, with capabilities available in both Starburst Galaxy and Starburst Enterprise Platform.