Virtual Agents
on 01-23-2025 12:00 AM by SnapApp by BlueVector AI
81
Virtual Agents are AI-powered assistants that can be configured within SnapApp. They provide intelligent assistance by leveraging custom knowledge bases and various data sources. Virtual Agents can understand user queries, process information, and execute complex workflows based on the configured settings and available data sources.
Table of Contents
Create a Virtual Agent
- Navigate to Settings from the User menu of the top bar.
- Tap on Automation of the left navigation bar → Virtual Agents.
- Tap on +Add New in the top header of the Actions list view.
- Configure the Virtual Agent settings and fill out the required fields.
| Fields | Description |
|---|---|
| Name | Unique identifier for the Virtual Agent. |
| Display Options | Determines where the Virtual Agent will appear (Page, Popup Window, Side Panel). |
| Description | Provides a clear explanation of the Virtual Agent’s purpose for the user. |
| Goal | A goal is a high level description of what the virtual agent should accomplish. |
| Workflow Prompt | Outline the responsibilities of the Virtual Agent (VA) using bullet points. Clearly define all functionalities, scenarios, conditions, and decision paths for the VA. Specify what the VA will do in various cases and what information it will provide to users. Use placeholders in curly brackets {} denote specific pieces of information the VA can reference. These placeholders are referred to as Groundings. |
For example:
Groundings
GROUNDINGS are the data sources that Virtual Agents use to evaluate and respond to queries. These data sources provide the necessary context for generating accurate and relevant responses.
| Field | Description |
|---|---|
| Name | Unique name identifying the grounding. |
| Type | Specifies the type of grounding. Below are the available types: |
| 1. Datastore: Google Vertex AI embedding store used for storing vector data that models utilize. | |
| 2. Expression: Executes specific expressions based on user queries and returns the result. | |
| 3. Workflow: Executes workflows. It could be used for actions like sending emails , field updates etc. | |
| 4. SQL GenSearch: Allows direct querying of object data using SnapApp generative search. | |
| 5. Application Breadcrumbs: (Feature not implemented yet). | |
| Object | Select the existing object for which the grounding is created. |
| Description | Include detailed descriptions to help the Virtual Agent understand when and how to use a grounding. |
| Status | Activates or deactivates the grounding. Options: Active or Inactive (default). |
| Preloading Responses | A list of predefined messages displayed while the system generates a response, ensuring the user remains engaged. |
Note: Preloading Responses — It is in a list format. You can add multiple preloading responses, and a random one will be used each time.
Example: LIST("hold on", "loading", "fetching you info")
Note: When the type of grounding is “expression,” ensure that the description explains the input parameters required for the expression. Provide clear and precise explanations for each parameter within the grounding’s description.
Thank you for following these steps to configure your SnapApp components effectively If you have any questions or need further assistance, please don’t hesitate to reach out to our support team. We’re here to help you make the most out of your SnapApp experience.
For support, email us at snapapp@bluevector.ai
