Back to Browse

Run LLMs in SQL: Databricks ai_query() for Sentiment Analysis & Summarization

295 views
Oct 10, 2025
12:57

Tired of complex pipelines just to run a quick sentiment analysis? You don't need Python or Spark for LLMs anymore!This video shows you how to use the game-changing Databricks ai_query() SQL function to run Large Language Models (LLMs) directly on your data. We walk through a full example—from creating a table to performing real-time sentiment analysis and text summarization on customer reviews, all within the Databricks SQL editor. What you will learn: How to create a dynamic prompt for the ai_query() function. Using the databricks-gpt-oss-20b model for batch inference. Extracting structured data (like 'Positive' or 'Negative' sentiment) from unstructured text. The power of bringing Generative AI into your Lakehouse at scale. SQL Code Used in Video Documentation: https://deltafrog.net/ai_query_databricks/ #databrickstutorial #LLM #SQL #ai_query #SentimentAnalysis #DataEngineering #GenerativeAI

Download

1 formats

Video Formats

360pmp418.0 MB

Right-click 'Download' and select 'Save Link As' if the file opens in a new tab.

Run LLMs in SQL: Databricks ai_query() for Sentiment Analysis & Summarization | NatokHD