SnapGPT settings

SnapGPT uses Large Language Models (LLMs) to respond to text prompts. SnapGPT can:

Generate pipelines
Document existing pipelines
Write expressions to use in Snaps
Ask questions about the SnapLogic Platform

Refer to SnapGPT - Security and Data Handling Protocols for information on how SnapGPT works and the SnapGPT documentation for examples on how to use it.

The SnapGPT settings screen is available only in Admin Manager and provides the following options:

Settings

The following SnapGPT settings are available for your Org.


Setting		Purpose
Enable SnapGPT		Turns SnapGPT on for your Org.
Schema retrieval		When enabled, SnapGPT infers the schema to create more accurate SQL queries in SQL Snaps. If disabled, SnapGPT still creates SQL queries, but users need edit those for the relevant schema.
Use preview data schema for expression generation		Enable to preview of data schema to write expressions. If disabled, SnapGPT-generated expressions might have lower accuracy.
Describe pipeline		Enables pipeline documentation without sending any sensitive information to the LLM.
Use my existing pipelines for enhanced pipeline generation	`Auto`	SnapGPT automatically selects pipelines to be used in Retrieval Augmented Generation. This feature selects pipelines in your environment to use as examples for generating new pipelines.
	`Manual`	Enables Retrieval Augmented Generation for user to select pipelines manually.
	`Disable`	Disables Retrieval Augmented Generation.

Attention: In the current release, SnapGPT cannot read schemas. The settings for configuring schema access will apply when that functionality is available.