SnapGPT

Overview

SnapGPT leverages the power of generative AI to enable you to understand and build SnapLogic pipelines.


SnapGPT panel in Designer

Features

You can use SnapGPT to:
  • Build pipelines using simple prompts
  • Describe pipelines
  • Analyze pipelines
  • Ask anything about the SnapLogic Intelligent Integration Platform (IIP)
  • Generate SnapLogic expressions
  • Create SQL queries

Skills

SnapGPT Skills offer interactive flows to guide you toward the outcome you want to achieve.

SnapGPT provides these three Skills.
  • Pipeline Skills: Enable you to generate pipelines based on your instructions. We highly encourage you to experiment earnestly with these flows, which provide a more self-guided experience. You can choose from any one of the following flows:
    • Generate a pipeline through prompts. You can choose to implement canonical pipelines from the Public Pattern Library for additional context. When the pipeline is generated, you can proceed with learning more about that skill.
    • Get an analysis of the pipeline you have on the Designer canvas. The analysis reports errors and other issues with the pipeline design. This flow helps with the iterative development of a pipeline.
    • View a description of a pipeline. You can ask SnapGPT to describe a pipeline through an upload or by pasting the .slp file contents into the window. The response includes pipeline notes, enabling a communication loop for the user and pipeline designer. This flow might prove fruitful for those who learn by example. You can browse pipelines in the Pattern or Pipeline Catalogs and analyze them.
  • Contextual Skills offer learning guidance within the context of Snap configuration activities. You can view video content and documentation to understand how to generate an expression, configure the Mapper Snap, and build SQL statements.
  • Knowledge Skills give initial guidance through suggestions. This flow enables you to get started with SnapGPT.

To improve interactions, SnapGPT provides links in all flows that reference other learning resources.

SnapGPT is powered by publicly-available large language models (LLMs), so the prompts you enter query the entire data set of the LLMs and is not limited to specific SnapLogic data. Generic parts of prompts are sent anonymously to the supported LLMs.

Controls

  • Launch the SnapGPT app by click SnapGPT at the top of Designer.

  • At the top of the SnapGPT pane is a toolbar:

    You can perform the following actions by clicking the respective icons from left to right:

    • Maximize the SnapGPT Pane across the Designer workbench.
    • Return the SnapGPT display back to the Skills menu. This is useful if you're scrolling through suggestions.
    • Open SnapGPT documentation.
    • Open the User Settings dialog.
    • Close the SnapGPT pane.
  • Additionally, when SnapGPT is open, detach the panel and move it anywhere on the Designer canvas. You can drag and resize the panel to see chat content clearer.

SnapGPT Settings

SnapGPT Settings control the methods for generating responses in a user session. You can modify the settings per session. These settings differ from the Environment (Org) settings for SnapGPT.

Click the gear icon to open the Settings dialog. You can opt to have SnapGPT open every time you navigate to Designer. For each SnapGPT session, you can do the following:
  1. Select the LLM from the dropdown. The list populates with the LLMs in your Org/Environment.
  2. Select the level of multiple-mode LLM requests.
  3. Enter a list of pipeline IDs for reference to template pipelines for better contexts.
  4. Select the Retrieval Augmented Generation (RAG) settings (Auto, Manual, and Disabled).

Renewal Augmented Generation (RAG)

RAG is a way to improve results based on existing assets in your environment (Org). Its implementation in the SnapLogic platform helps power accurate results based on what the you are asking for. Because LLMs do not have context before 2022, RAG adds those extra contexts from your knowledge archive or personal databases. RAG also decreases the chance of LLM hallucinations when the most accurate answer is unavailable.

RAG settings are set by the Org admin in Admin Manager > SnapGPT Settings. The Org settings determine the settings available to the user.

Users can also specify the RAG settings in their interaction with SnapGPT in the Select patterns or pipelines to improve quality dialog. Refer to Select Examples to Generate Pipelines for an example workflow.

RAG settings matrix

The following table describes the relationship between Org and user settings:

Environment Setting User Setting Result
Auto Auto

When Auto is selected, pipelines are recommended by SnapGPT. Suggests 5 pipelines for you to choose from, to generate a new pipeline.
Manual or Disable You can also choose either to select the recommended pipelines manually or to disable the RAG settings entirely.
Manual Manual or Disable You can also choose either to select the recommended pipelines manually or to disable the RAG settings entirely.
Disable N/A RAG is disabled.

Known issues

  • You might experience significant latency (up to one minute) or timeout issues, depending on the complexity of the request and the capacity of the LLM provider.
  • As with all large language models, SnapGPT can occasionally be unreliable or inconsistent in its responses.