Lucidworks AI

Table of Contents

Pre-trained embedding models
Custom embedding model training
Seamless integration with Fusion

This feature is currently only available to clients who have contracted with Lucidworks for features related to Neural Hybrid Search and Lucidworks AI.

Lucidworks AI provides easy-to-use, AI-powered data discovery and search capabilities including:

Pre-trained embedding models
Custom embedding model training
Seamless integration with Fusion

In order to use Lucidworks AI, you must set up the AI Gateway.

Pre-trained embedding models

Lucidworks provides state-of-the-art pre-trained embedding models to support semantic vector search (SVS).

You can use Lucidworks AI to choose the models that best suit your organizational needs.
Lucidworks consistently evaluates and tests new models and enhancements to existing models. When an enhancement or a new model is verified to be effective on the platform, it will be released.

Custom embedding model training

Lucidworks provides a user-friendly interface that lets you quickly train, deploy, and manage custom models.

Custom trained models are essential for eCommerce and other use cases with a domain-specific language.
Custom model training uses:
- Signals data that enables optimization for particular business use cases and metrics.
- Graphics processing units (GPUs) for optimal performance.
The Lucidworks AI model training user interface provides evaluation metrics to easily compare trained models.
Lucidworks provides small and efficient custom models that scale to thousands of queries per second (QPS) without decreasing the quality of embeddings.

Seamless integration with Fusion

To utilize models within a Fusion pipeline, Lucidworks provides query and indexing pipeline stages that integrate with Lucidworks AI.