This guide walks you through creating your first AI agent in Exei and configuring its core setup.
In Exei, AI agents are the foundation of the platform. Each agent is created for a specific role, trained on your data, and deployed across one or more channels to interact with users.
This guide applies to all agent types, including:
Customer Service agent
Employee Support agents
Shopping Assistant agents
After signing in, you are redirected to the My Agents page.
The My Agents page is the central place where you can:
View all AI agents created under your organization.
Create new AI agents.
Open and manage existing agents.
If no agents exist yet, you will see an option to create your first AI agent.
To create a new AI agent:
Go to the My Agents page.
Click Create AI Agent.
Fill in the required details.
Save and proceed to agent setup.
Once created, the agent will appear in your agents list and can be configured further.
When creating an AI agent, you must provide the following details.
The Agent Name is the display name of your AI agent.
Choose a name wisely as the AI Agent will be aware of what you are naming it, such as:
Digi AI
Ashish
Kyra
This name is visible to users during conversations.
The Agent Role defines the primary function of the AI agent.
Exei supports three roles:
Customer Service – For handling customer support queries.
Employee Support – For internal employee assistance.
Shopping Assistant – For product related queries and recommendations
The selected role determines default behavior and available features.
Once the agent role is selected, you configure the AI persona.This is a direct instruction to your Agent. These settings define how the agent behaves and communicates.
This is the name of your organization as it should appear in agent responses.
It helps the agent maintain brand consistency while interacting with users.
Select the industry that best represents your business. This helps the agent respond with appropriate context and terminology.
Your Organization's industry getts pre-filled, you can change it for the Agent.
Purpose defines what the agent is expected to do.
Exei provides predefined purposes based on the selected agent role. You can:
Use the suggested purpose as-is.
Modify it to better match your use case.
Write a fully custom purpose.
Clear purposes lead to more accurate responses.
Personality traits define how the agent behaves during conversations.
Examples include:
Helpful
Professional
Friendly
Detail-oriented
Concise
Supportive
You can select multiple traits to shape the agent’s communication style.
Tone of Voice controls how formal or casual the agent sounds.
Available options are:
Professional
Friendly
Casual
Choose a tone that aligns with your brand and audience.
While agent creation is unified, behavior differs slightly by role.
Typically used for:
Answering FAQs
Handling support queries
Creating service tickets
Escalating to human agents
These agents are commonly deployed on websites, WhatsApp, and voice channels.
Typically used for:
Internal policies and documentation
HR or IT-related questions
Employee workflows and requests
These agents are usually deployed internally.
Designed for ecommerce use cases across different platforms, including Shopify and custom websites.
Shopping Assistant agents can:
Assist users during product discovery and purchase decisions.
Answer shopping and product-related questions.
Support users throughout the buying journey.
Handle common purchase-related queries.
These agents are optimized for ecommerce interactions and can be deployed on storefronts or websites, with product catalogs and recommendation logic configured separately.
Once the agent is created, you can proceed to:
Train the agent using the Knowledge Base.
Add products (for Shopping Assistant agents).
Configure channels and deploy the agent.
Enable AI Actions such as leads, appointments, or tickets.
Test the agent before going live.
Agent creation is only the first step; training and deployment complete the setup.
Start with a clear and focused purpose..
Choose personality traits that match your brand
Keep tone consistent across agents.
Create separate agents for different use cases.
Test the agent before deploying to users.
Now that your AI agent is created, the next step is to train it with your data.
Next User Guide: Agent Customization