Milvus VDB Provider

The Milvus VDB Provider module enables Vector Database (VDB) support for AI-driven functionality in Drupal. It integrates with the AI Core module and AI Search module to perform high-performance vector searches using embeddings, making it ideal for semantic search, content recommendations, and AI-powered search experiences.

Key Features

• Vector search integration

Seamlessly connects your Drupal site's AI module to a vector database backend, enabling similarity-based searches using embeddings. Supports multiple similarity metrics including cosine similarity, Euclidean distance (L2), and inner product (IP) for flexible search strategies.

• Self-host or cloud deployment

Use your own open-source Milvus server (including DDEV integration for local development), or opt for Zilliz Cloud if you prefer a managed, hosted solution with no infrastructure management.

• Flexible storage and CRUD operations

Complete support for creating, reading, updating, and deleting vector data with metadata tagging, dynamic fields, and automatic indexing for fast retrieval. Collections are automatically created and managed based on your Search API configuration.

• Advanced filtering and grouping

Build complex queries with advanced filtering capabilities and result grouping by entity ID. Supports nested condition groups and multiple field types for precise search control.

• Scalable architecture

Built to handle large-scale vector datasets, taking advantage of Milvus' efficient indexing, high throughput, and distributed architecture for production-ready performance.

• Secure connections

Supports API key and username/password authentication for secure communication with both Milvus (self-hosted) and Zilliz Cloud deployments, with full integration with Drupal's Key module for credential management.