LLM Vector Store Base for Odoo

This module extends the base LLM integration module to provide support for vector databases in Odoo applications. It offers a framework for connecting to various vector store providers and abstracting away the implementation details, making it easy to build RAG (Retrieval Augmented Generation) applications.

Features

• Vector Store Connectivity: Connect to various vector databases
• Provider-Agnostic Interface: Common API across different vector store providers
• Collection Management: Abstract model for working with vector collections
• Vector Operations: Insert, search, and delete vectors with metadata

Vector Store Framework

The vector store framework allows you to:

1. Connect to Various Vector Databases: Support for vector stores like Qdrant, Chroma, PostgreSQL, and others through extension modules
2. Manage Collections: Create, delete, and list collections within your vector store
3. Vector Operations: Insert, search, and delete vectors with metadata
4. Index Management: Create and manage indices for optimal search performance

Architecture

The vector store integration follows a provider pattern:

• Base Models: llm.store provides the foundation
• Abstract Collection Model: llm.store.collection serves as a base for concrete implementations
• Provider Pattern: Extensible architecture using dispatch methods to support different vector store implementations
• No Implementation Lock-in: Easily switch between vector store providers with a consistent API

Usage Example

# Get a configured vector store
store = env.ref('your_module.your_vector_store')

# Create a collection
store.create_collection('my_collection', dimension=1536)

# Insert vectors
vectors = [[0.1, 0.2, ...], [0.3, 0.4, ...]]  # Your embedding vectors
metadata = [{'text': 'Document 1'}, {'text': 'Document 2'}]
store.insert_vectors('my_collection', vectors, metadata=metadata)

# Search vectors
query_vector = [0.2, 0.3, ...]  # Your query embedding
results = store.search_vectors('my_collection', query_vector, limit=5)

Installation

1. Ensure the base LLM module is installed
2. Install this module in your Odoo instance
3. Configure vector stores through the LLM > Configuration > Vector Stores menu

Extending with Providers

To add support for a specific vector database:

1. Create a new module that depends on llm_store
2. Extend the llm.store model and implement the service-specific methods
3. Register your service in the selection field through _get_available_services()

Example:

class MyVectorStore(models.Model):
    _inherit = "llm.store"
    
    @api.model
    def _get_available_services(self):
        return super()._get_available_services() + [('my_provider', 'My Vector Store')]
    
    def my_provider_create_collection(self, name, dimension=None, metadata=None, **kwargs):
        # Implementation for creating a collection in your vector store
        pass
        
    # Implement other methods...

Security

The module uses the same security model as the base LLM module, with the LLM Manager group having full access to vector stores.

License

LGPL-3