Google Gemini Module

Updated on 31 Dec 2024
2 Minutes to read

Print
Share
Dark
Light

Article summary

Did you find this summary helpful?

Thank you for your feedback

Overview

Module Details
Core or Github Module	Core
Restart Required	No
Step Location	AI > Google Gemini
Settings Location	Settings > Google Gemini Settings
Prerequisites	A secret key must be configured in Settings prior to use. The Module must be added as a Project Dependency. Dependent Module(s): AI Common

The Google Gemini Module enables Users to access Google Gemini AI through the built in Chat Completion step.

Once prerequisites are installed, the Module automatically exposes the Chat Completion step in the Toolbox. This step includes a properties panel with fields that can be configured according to user preference.

Google Gemini works directly with both the OpenAI and AI Anthropic Modules, allowing Users to switch between Modules efficiently without reformatting prompts or replacing Steps.

Configuration/Properties

Installation

To set up the Module in Decisions, navigate to Settings > Administration > Features, locate the Google Gemini Module and select Install.
- The AI Common Module will automatically install in the background if this is the first AI Module installed in the environment.

Creating a Dependency:

After installing the module, navigate to the project that will utilize Google Gemini.
- Then within the Project, navigate to Manage > Configuration > Dependencies.
- Select Manage, navigate to Modules and select Decisions.AI.GoogleGemini.
- Select Confirm to save the change.

For more information on installing Modules please visit Installing Modules in Decisions.

For more information about creating a project dependency for a Module please visit Adding Modules as a Project Dependency

Available Steps

The table below provides a list of all the steps within the module.

Step Name	Description
Chat Completion	The Chat Completion step enables prompts to be submitted to a large language model(LLM) and will return the response the LLM gives back. Prompts can be added directly or pulled from the Flow. Users can also select the model they want to review the prompt.

Use Case: Utilizing Chat Completion Step with Google Gemini AI

To utilize Google Gemini's Chat Completion step in Decisions:

Drag and drop the Chat Completion step onto the workspace and select the Step.
Navigate to the Properties Panel on the right and check the Override Default Provider box.
- Select Decisions.AI.GoogleGemini from the drop-down menu.
Next, check the Select Model box and choose the desired Google Gemini Model from the drop-down menu. Override and either create a constant prompt directly from a Flow or utilize a pre constructed prompt made from the AI Prompt Manager.
Users have two options when selecting an AI Prompt:
- To create a prompt within the Flow, check Get Prompt from Flow.
- To utilize a prompt created in the AI Prompt Manager, click Select Prompt.

To add Find and Replace Variables, select Build Array on Override Variables and Build Data on Item 0.
- Add desired variables in both the Find and Replace fields that correspond with the AI Prompt.

Once the desired settings have been saved, Select Debug.
- The Debugger should display the Chat Response that includes settings customized in the Properties Panel.