What to Install

This module is typically auto-installed as a dependency. You rarely need to install it directly.

For RAG/Knowledge Base features:

# Install knowledge module with a vector store
odoo-bin -d your_db -i llm_knowledge,llm_pgvector

Auto-Installed Dependencies

These are pulled in automatically:

• llm (core infrastructure)
• mail (Odoo messaging)

Choose a Vector Store Implementation

Common Setups

Core Capabilities

• Multi-Provider Support - Unified interface for ChromaDB, pgvector, Qdrant, and other vector stores
• Collection Management - Abstract models for organizing and managing vector collections
• Vector Operations - Insert, search, update, and delete operations with metadata support
• Index Management - Automatic index creation and optimization for performance
• RAG Integration - Seamless integration with knowledge base and retrieval systems

Provider Abstraction Framework

Unified Interface Across Vector Stores:

class LLMStore(models.Model):
    _name = "llm.store"
    _description = "LLM Vector Store"

    def _dispatch(self, method, *args, **kwargs):
        """Dynamic dispatch to service-specific implementation"""
        service_method = f"{self.service}_{method}"
        if hasattr(self, service_method):
            return getattr(self, service_method)(*args, **kwargs)

Collection Management

# Create collection
collection = env['llm.store.collection'].create({
    'name': 'knowledge_base',
    'store_id': store.id,
    'dimension': 1536,  # OpenAI embedding dimension
    'distance_metric': 'cosine'
})
collection.create_collection()

# Insert vectors
collection.insert_vectors(
    vectors=[[0.1, 0.2, 0.3, ...]],
    metadata=[{'document_id': 123}],
    ids=['doc_123']
)

# Search vectors
results = collection.search_vectors(
    query_vector=[0.2, 0.3, 0.4, ...],
    limit=5
)

Core Store Methods

# Collection management
def create_collection(self, name, dimension, metric='cosine'):
    """Create new vector collection"""

def delete_collection(self, name):
    """Delete vector collection"""

# Vector operations
def insert_vectors(self, collection, vectors, metadata=None, ids=None):
    """Insert vectors into collection"""

def search_vectors(self, collection, query_vector, limit=10, filter=None):
    """Search similar vectors"""

def update_vectors(self, collection, ids, vectors=None, metadata=None):
    """Update existing vectors"""

def delete_vectors(self, collection, ids=None, filter=None):
    """Delete vectors from collection"""

Module Information

• Name: LLM Vector Store Base
• Version: 18.0.1.0.0
• Category: Technical
• License: LGPL-3
• Dependencies: llm, mail
• Author: Apexive Solutions LLC

Key Models

• ``llm.store``: Base vector store configuration
• ``llm.store.collection``: Vector collection management

Performance Characteristics