Quickstart: Helpdesk Agent

Get an AI agent handling your helpdesk tickets in about 10 minutes

This guide walks you through connecting your helpdesk and getting an AI agent that responds to customer tickets. By the end, you'll have an agent trained on your help center, ready to handle incoming tickets.

Time: About 10 minutes You'll need: An account on one of our supported helpdesks (Zendesk, Freshdesk, Intercom, Gorgias, or Jira)

Step 1: Create your agent

1. Sign in to dashboard.eesel.ai

2. Click Create Agent and select Help Desk Agent

3. Give your agent a name (e.g., "Support Agent" or "Tier 1 Agent")

Step 2: Connect your helpdesk

1. In your agent workspace, go to Integrations

2. Select your helpdesk:

   • Zendesk

   • Freshdesk

   • Intercom

   • Gorgias

   • Jira

3. Follow the connection steps for your helpdesk

Quick start option: Most helpdesks let you start by just entering your help center URL. This trains your agent on your public articles without needing to connect your full account yet.

Step 3: Train on your knowledge

Once connected, your agent automatically starts learning from your help center articles. You can also add:

• Past tickets — Select "Tickets" in your helpdesk integration to train on your team's best replies

• Macros/Canned responses — Import your existing macro library so the agent uses consistent responses

• Other sources — Add Google Drive docs, Confluence pages, PDFs, or your website

Training typically takes a few minutes depending on how much content you have.

Step 4: Test your agent

1. Use the test chat in your agent workspace to ask questions

2. Try questions your customers commonly ask

3. Check that responses are accurate and sound right

Tips for testing:

• Ask the same questions you see in your ticket queue

• Test edge cases — what happens when the agent doesn't know the answer?

• Check that it escalates properly when it should

Step 5: Configure actions

Set up what your agent can do beyond just answering questions:

Action	What it does
Leave public reply	Send a response directly to the customer
Leave draft reply	Write a draft for your team to review first
Tag ticket	Automatically categorize tickets
Assign ticket	Route to the right team member or group
Close ticket	Close tickets after resolving them
Leave internal note	Add notes only your team can see

We recommend starting with Leave draft reply so you can review responses before they go to customers. Once you're confident, switch to Leave public reply for full automation.

Step 6: Set up a trigger

Triggers tell your agent when to activate. Go to Settings > Triggers and choose:

• Any Customer Message — Agent responds to every incoming ticket

• First Customer Message — Agent only handles the first message (great for triage)

• Agent Requests Help — Agent activates when your team @mentions it in internal notes

Step 7: Go live

1. Enable your trigger to start handling real tickets

2. Consider turning on Human in the Loop for the first few days to review responses before they send

3. Monitor responses in your helpdesk and fine-tune the instructions as needed

You're live! Your AI agent is now handling tickets. Check Analytics to track performance.

What's next?

• Write better instructions to fine-tune responses

• Add more knowledge sources to improve accuracy

• Learn about AI Copilot for agent-assist mode

• Set up AI Triage to auto-categorize and route tickets