AI Agents feature overview

Home > Technical Support > Features and functions > AI > AI features > AI agents > AI Agents feature overview

The AI Agents feature allows you to create reusable AI-powered agents and trigger them via automatic rules. AI agents utilize the built-in MCP server and its tools to perform various actions within your LiveAgent account. They serve as a background automation layer alongside other LiveAgent AI-powered features.

Note: To use the AI Agents feature, it is required to integrate your LiveAgent account with an AI provider, and set the feature up.

It is recommended to integrate FlowHunt as your AI provider and set up your AI agents via it to have full control over the workflow and troubleshooting.

FlowHunt, our no-code AI automation platform, is a product of our company, Quality Unit, like LiveAgent. By utilizing our AI features with FlowHunt, we ensure data safety and offer priority support. For customers with active LiveAgent & FlowHunt subscriptions, our support team offers a full initial setup. If you're interested in our free setup service, feel free to contact us at support@liveagent.com.

How it works

An AI agent is the central configuration object. It tells the system which AI provider to call, how it should process ticket(s), and which actions it is allowed to perform. One agent per use case (e.g., one for auto-tagging, one for ticket triage/categorization, one for generating knowledge base articles from resolved tickets, etc.) is the recommended approach. One AI agent can be reused in many automatic rules, and each AI agent has its own AI provider and model.

When an automatic rule containing the action "Run AI agent" triggers, the system dispatches the AI agent to your AI provider, which processes the ticket and performs actions it is allowed to do, such as tagging the ticket, resolving it, assigning it to an agent, transferring it to another department, or writing a knowledge base article out of it - all without any human intervention.

LiveAgent exposes a built-in MCP server (model context protocol) that your AI provider connects to when executing an AI agent. When the AI provider needs to read ticket data or perform an action, it calls an MCP server endpoint, and LiveAgent executes the corresponding tool. Every tool call is authenticated and validated before it reaches the ticket.

Creating an AI agent

To create & set up an AI agent, navigate to Configuration > AI > AI Agents. The exact setup steps depend on the AI provider you use. Please follow the dedicated article for a step-by-step setup guide for your provider:

Adjusting an AI agent's configuration

To adjust the configuration of an existing AI agent, navigate again to Configuration > AI > AI Agents, and click on the desired AI agent. You can change the provider, model (or FlowHunt workflow), instructions, allowed tools, and access token of your AI agent anytime as required.

AI agent tools

Each AI agent includes a list of allowed tools - specific actions it may invoke when processing a ticket. Tools that are not allowed for a given AI agent are unavailable to the AI provider, even if the underlying provider would request them.

Note: See the MCP server tools list for the complete list of all available tools and their description.

Tip: It is recommended to grant each AI agent only the tools it actually needs. For example, an agent whose sole purpose is to tag tickets does not need the "Resolve ticket" or "Assign ticket" tools. Keeping the tool list tight reduces the risk of unintended actions and helps keep tool calls, and therefore your AI budget usage, low.

AI budgets for AI agents

AI budgets are a way to limit how much your AI agents are used. A budget limits the number of MCP tool calls that the AI agents tied to it can make in a month. For example, an AI agent run that uses 3 MCP tools to get ticket messages, get ticket notes, and create a knowledge base article makes 3 tool calls.

Each run is tied to a budget you choose when setting up the rule action. For a full explanation of how AI budgets work, please see the separate AI budgets article.

Note: AI budgets are not a price. LiveAgent does not charge for AI. The cost of running AI agents is billed by your AI provider, based on your usage there.

Tip: Keep an eye on your tool call usage and design your AI agents to use tool calls efficiently, so that your budgets last and your AI provider usage stays low.

Monitoring AI agent runs

Every time an AI agent is triggered for a ticket, a run record that tracks its full lifecycle is created in the Reports > AI Agent Runs section. Each run record tracks the status, the number of tool calls it made, and, if it failed, the reason. For a more detailed explanation of the AI Agent Runs report section, please see the AI Agent Runs report article.

Related articles

Updated:

Other Articles