Skip to main content

Uniconnect AI Agents Documentation

Overview

The Uniconnect AI Agents Management allows you to create, configure, and manage AI agents tailored to your workflows. Agents can be customized with roles, execution modes, and knowledge bases, enabling them to perform tasks, chat, or interact via voice.

This guide explains:

  • How to create an AI agent
  • The different agent types and their functions
  • How to connect agents to knowledge bases
  • Best practices for configuration

Creating an AI Agent

Step 1: Access Agent Management

Navigate to Agent Management in your Uniconnect dashboard. You’ll see two main panels:

  • Agents List (left): Displays all created agents. Initially, this shows No data until agents are added.
  • Configuration Panel (right): Contains tabs for General and Knowledge Base settings.

Step 2: Fill in General Information

Under the General tab, complete the following fields:

FieldDescription
AI Assistant NameEnter a unique name for your agent.
RoleAssign a role (e.g., Support, Sales, Workflow Automation).
GroupsSelect groups the agent belongs to (e.g., department or project teams).
Execution ModeChoose how the agent operates (see below).
LLM ProviderSelect the language model provider (e.g., OpenAI, Azure AI).
ModelChoose the specific model after selecting a provider.
LanguageDefine the agent’s default communication language.

Click Create (top right) once all fields are completed.


Execution Modes

Agents can operate in different modes depending on your needs:

ModeFunctionBest Use Case
Task AgentExecutes a single task and returns results.Workflow automation, quick actions.
Chat AgentEngages in conversational interaction, guiding tasks and answering queries.Customer support, knowledge assistance.
Voice AgentEnables voice-based interaction for hands-free communication.Call centers, accessibility support.

Connecting Knowledge Bases

Step 1: Open Knowledge Base Tab

Switch to the Knowledge Base tab in the configuration panel.

Step 2: Select Knowledge Sources

You’ll see a list of available knowledge bases with checkboxes, such as:

  • Attachments
  • Comments
  • Contacts etc.

Use the search bar to quickly find relevant sources. Pagination at the bottom allows browsing through multiple pages of knowledge bases.

Step 3: Save Configuration

After selecting the required knowledge bases, click Save (bottom right). You can also Cancel to discard changes.


Example Scenarios

  • Customer Support Agent (Chat Agent)

    • Role: Support
    • Groups: Customer Service
    • Knowledge Base: Attachments + CRM
    • Function: Answers customer queries with contextual data.
  • Workflow Automation Agent (Task Agent)

    • Role: Operations
    • Groups: Automation Team
    • Knowledge Base: Internal Documents
    • Function: Executes repetitive tasks like report generation.
  • Voice Assistant (Voice Agent)

    • Role: Sales
    • Groups: Field Team
    • Knowledge Base: Contacts + CRM
    • Function: Provides voice-driven access to client information.

Best Practices

  • Name agents clearly to reflect their purpose.
  • Assign roles and groups to align with organizational structure.
  • Choose execution mode wisely based on interaction type.
  • Link relevant knowledge bases for accurate responses.
  • Test agents after creation to ensure proper functionality.