Skip to content

[Integration]: Moss as Datasource #12414

@CoderOMaster

Description

@CoderOMaster

Is there an existing integration?

  • I have searched the existing integrations.

Use Case

Moss is a semantic search runtime built for Conversational AI agents. It lets you index documents and query them in under 10ms fast enough for real-time conversation, compared to 200–300ms with typical retrieval systems.

The MindsDB integration would allow users to connect a Moss instance as a data source, enabling semantic search queries directly within MindsDB's SQL interface and AI pipelines.

Motivation

Retrieval-Augmented Generation (RAG) pipelines are latency-sensitive slow retrieval breaks the conversational experience. Moss is purpose-built to solve this, but currently requires custom integration work in each project. A MindsDB handler would let users plug Moss into AI workflows without writing any custom retrieval code, and pair it with MindsDB's model layer for end-to-end RAG in pure SQL.

Implementation

I am looking into other existing integrations and i think final implementation of usage would look like:

CREATE DATABASE moss_datasource
WITH ENGINE = 'moss'
PARAMETERS = {
"index_name" : "sample_index",
"alpha": "0" -- optional, 0-1 range
}

Anything else?

Look up us on https://moss.dev

Metadata

Metadata

Assignees

No one assigned

    Labels

    integrationNew idea for the supported db integrations

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions