---
name: weaviate-integration
description: Weaviate vector database setup with GraphQL queries and hybrid search
allowed-tools:
  - Read
  - Write
  - Edit
  - Bash
  - Glob
  - Grep
graph:
  domains: [domain:software-engineering]
  specializations: [specialization:ai-agents-conversational]
  skillAreas: [skill-area:retrieval-augmented-generation, skill-area:search-indexing]
  roles: [role:ml-engineer, role:backend-engineer]
  workflows: [workflow:ml-model-lifecycle, workflow:feature-development]

---

# Weaviate Integration Skill

## Capabilities

- Set up Weaviate cluster (cloud or self-hosted)
- Define schemas with properties and vectorizers
- Implement GraphQL queries
- Configure hybrid search (vector + keyword)
- Set up multi-tenancy
- Implement batch import operations

## Target Processes

- vector-database-setup
- rag-pipeline-implementation

## Implementation Details

### Core Operations

1. **Schema Management**: Class definitions and properties
2. **Data Import**: Single and batch object creation
3. **Vector Search**: nearVector, nearText queries
4. **Hybrid Search**: Combined vector and BM25
5. **GraphQL**: Flexible querying with Get and Aggregate

### Configuration Options

- Vectorizer modules (text2vec-*, multi2vec-*)
- Replication factor
- Sharding configuration
- Multi-tenancy settings
- Module configuration

### Best Practices

- Design schema for query patterns
- Use appropriate vectorizer
- Enable hybrid search for better recall
- Configure proper backups
- Monitor resource usage

### Dependencies

- weaviate-client
- langchain-weaviate