databricks-vector-search

Create and query Databricks Vector Search indexes for semantic and hybrid retrieval.

1|Updated Apr 24, 2026
One-click install
npx skills add https://github.com/itsadijmbt/SecureMCP-Servers --skill databricks-vector-search-itsadijmbt
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: databricks-vector-search
Source: https://github.com/itsadijmbt/SecureMCP-Servers/tree/main/TEST_SERVERS/PORTED_TO_SECUREMCP/databrickslab-mcp/ai-dev-kit/databricks-skills/databricks-vector-search
Command: npx skills add https://github.com/itsadijmbt/SecureMCP-Servers --skill databricks-vector-search-itsadijmbt

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps you design vector search indexes and reliably query them for semantic and hybrid retrieval so your RAG or search app returns relevant results.

Core Features & Use Cases

  • Create and manage Vector Search endpoints and indexes: Spin up Standard or Storage-Optimized endpoints and configure Delta Sync (managed or self-managed embeddings) or Direct Access indexes.
  • Query for semantic and hybrid relevance: Run similarity search using query text, precomputed query vectors, optional hybrid mode, and endpoint-appropriate filters.
  • Keep indexes up to date: Use TRIGGERED pipeline sync for batch refreshes and monitor operational status for provisioning and indexing.

Quick Start

Use the databricks-vector-search skill to create a Storage-Optimized endpoint and a Delta Sync index for your documents, then query it by text with filters to retrieve the most relevant passages for a RAG app.

Frequently Asked Questions about databricks-vector-search

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a Databricks vector search index for RAG applications?

To build a Databricks vector search index for RAG, configure a Vector Search endpoint and create a Delta Sync index using managed or self-managed embeddings over your Delta tables. This enables semantic and hybrid retrieval for relevant passages.

What is the difference between Delta Sync and Direct Access vector indexes in Databricks?

Delta Sync indexes automatically synchronize and compute embeddings from source Delta tables, while Direct Access indexes require you to supply and manage precomputed vectors directly. Delta Sync supports both managed and self-managed embedding workflows.

How do I query a Databricks vector search index using text and filters?

Query a Databricks vector search index by passing query_text or a precomputed query_vector along with endpoint-appropriate filters. You can enable hybrid mode to combine semantic similarity with keyword filtering for improved relevance.

Why does my Databricks vector search index return no results or stay offline?

Your Databricks vector search index may stay offline if TRIGGERED pipelines have not been synchronized. Run sync operations to ensure the index status is ONLINE and queryable, and verify that your endpoint is fully provisioned before searching.

When should I use Standard vs Storage-Optimized endpoints for Databricks vector search?

Choose Standard or Storage-Optimized endpoints for Databricks vector search based on your scale and latency requirements. Storage-Optimized endpoints handle larger datasets, while Standard endpoints suit general-purpose semantic search and RAG workloads.