What problem does it solve?
This Skill solves the challenge of implementing complex, production-grade Databricks SQL workflows without sifting through scattered official documentation, covering everything from procedural scripting and geospatial analytics to AI-powered data enrichment and performance optimization.
Core Features & Use Cases
- Procedural SQL & Stored Procedures: Build reusable ETL pipelines with error handling, loops, and dynamic SQL for complex transformations, such as a customer upsert procedure that logs failures and returns row counts.
- Advanced Query Patterns: Use pipe syntax for readable, DataFrame-like query pipelines, recursive CTEs for hierarchical data traversal (e.g., organizational charts or bill of materials), and materialized views for precomputed, automatically refreshed aggregations.
- Specialized Built-in Functions: Leverage 39+ geospatial functions (H3 indexing, spatial joins), collation settings for case/accent-insensitive search, AI functions for text classification and summarization, http_request for external API calls, remote_query for federated database access, and read_files for direct volume ingestion.
- Performance & Modeling Best Practices: Optimize query speed with Liquid Clustering, caching, and proper dimensional modeling, while following anti-pattern avoidance guidelines to reduce compute costs and improve maintainability.
Quick Start
Use the databricks-dbsql skill to write a SQL query that classifies the 100 most recent support tickets into billing, technical, or account categories using the ai_classify function, and returns each ticket's ID, original description, and assigned category.
