Contrib.social

Add a trait, base class and interface for AI Search API backend engines

Open on Drupal.org →
Created on 30 May 2024, 6 months ago
Updated 13 June 2024, 5 months ago

Problem/Motivation

The backends that will be used for vector searching needs to be able to plugin to embeddings engines and needs to be able to configure this already on backend creation, since many of the databases focused on vectors, needs the vector storage to be defined already on creation.

Proposed resolution

Create a trait that holds all the needed functions to load the embeddings engine, add the embeddings engines extra configurations to the server backend form and to generate embeddings using the configuration set there.

Create a base class that uses this trait when setting up the form, so that any normal backend server using vectors can just implement the parent:: version of each function.

Create an interface that will be followed automatically by loading the trait. This is so that chatbots or other services can recognize the backend as a vector database.

✨ Feature request
Status

Fixed

Version

1.0

Component

Code

Created by

🇩🇪Germany marcus_johansson

Live updates comments and jobs are added and updated live.
Sign in to follow issues

Merge Requests

Comments & Activities

contrib.social Blog FAQ Discussions
Production build 0.71.5 2024