Agents Overview

AI Agents are the core of your automated conversation system, allowing you to create tailored processes for each lead source. These intelligent agents handle customer interactions, qualify leads, and guide prospects through your sales funnel with personalized, context-aware responses.

Purpose

Build and manage AI processes (“agents”).

Explanation

We use Agents to create a tailored agent for each lead source. Examples include:

Old Data
Facebook Leads
Website Leads
Credit Leads
Autotrader Leads
Cargurus Leads
Kijiji Auto Leads
Website Chatbot Leads
Service Department Leads

For each lead source, we start from a proven baseline agent and create a System Prompt tailored to that source.

What are AI Agents?

AI Agents are sophisticated conversation processes that automate customer interactions based on specific lead sources and business contexts. Each agent is equipped with a customized system prompt that defines its personality, knowledge, and conversation flow.

Each System Prompt should define:

Persona
Process (in-store, over-the-phone, hybrid)
Context on the specific lead source
Context on the type of client it may encounter
Context on the deal type (Auto, Powersport, RV, Hybrid)
Tasks (e.g., qualify the lead, book an appointment, get permission to pull credit)
Conversation flow
Direct question handling framework
Objection handling list
FAQ responses (approved)

Getting Started - New to agents? Learn the basics and create your first AI agent.
Agent Playground - Test and refine your agents in a safe testing environment.
Agent Prompts - Learn how to configure system prompts and agent behavior.
Auto Pilot Mode - Set up fully automated responses with intelligent timing.
Editing Agents - Modify and improve your existing agents effectively.
Troubleshooting - Solve common issues and get help when needed.
Best Practices - Follow proven guidelines for high-performing agents.

When to use Agents

Create/update processes that power conversations and workflows.

Prerequisites

Not all users have permission to edit pre-created agents. This safeguards optimized prompts and prevents breaking fine-tuned agents.

Creating an Agent

Any user can create an agent from scratch for any part of a structured process with an end goal.

How to create an Agent

Click ”+ Create New Process” on the Agents page. We recommend proper prompt engineering skills. For internal team members, we suggest starting from a pre-existing agent’s prompt.

Agent Playground

Selecting an agent brings you to the playground to quickly test the agent’s flow from a customer’s perspective. Start with an example opening message.

Example opening message the client would receive (Paid Lead Agent): “Hey Sam, this is Wes with Canada Drives. We just got your application for a new truck and would love to start showing you some options as soon as possible. Did you have a specific year, make, and model in mind?”

In the playground, respond like a customer would. Example chat input: “Yeah, I want a 2025 Dodge Ram 1500.”

This will kick off the agent’s conversation flow so you can test the flow, objections, and FAQ responses.

How Agents Work

Agents are text-based and built on an “AAI Text Model.” You can select an agent when:

Creating a new contact
Bulk SMS campaign
In the Conversations tab

The AI agent reads conversation history and its system prompt to respond appropriately.

Auto Pilot

Auto Pilot can be enabled for inbound leads or outbound SMS campaigns. It can be toggled on/off during a conversation. The agent automatically replies after a configured response time (45 seconds recommended).

Note: If you turn Auto Pilot OFF mid-conversation, the AI may still send that reply if it’s within the response window. If you turn Auto Pilot ON after an inbound message, click Generate Response once to re-engage, then leave Auto Pilot ON.

Generating an AI Response Manually

You can generate a response at any time from the chat input in Conversations. This allows reviewing and editing the AI’s reply before sending, which is useful for sensitive or high-stakes interactions while maintaining the agent’s flow.

Editing an Existing Agent

Steps:

On the Agents page, select the desired agent.
Once the agent is selected and inside the playground, select “Prompt editor” in the top right corner.
This will bring you to the code viewer.

Two ways to edit an agent:

In the code style.
In the right-hand side “what are we automating section” by typing input and requesting desired changes.

Code Editor

The code editor is the quickest and most effective way to edit a prompt. It’s similar to a document view but meticulously sectioned off using XML tags, which helps the agent understand the documentation.

Editing Instructions:

To edit a prompt (e.g., an objection handle like “whats my rate”), click in the code, use Command + F (Mac) or Control + F (PC) to search for the response ID “WhatsTheRate”, and then delete the text between the tags.

Example:

Scenario: Changing the objection handle for “whats the rate”.

Original Code Block:

<Response id="WhatsTheRate">
Great question, @firstName. Because of the amount of volume we do with the banks we get our
clients 2-3% lower than any other dealer and as low as 5.99% on used! Now your rate reflects
your current credit as well as the year, make, model, and mileage of the vehicle. We partner with
21 national lenders for exclusive rate breaks. I just need to build your file and send it to the banks
so we can get an accurate number for you. That way, we are not misleading you in any way. Sound
fair?
</Response>

Instruction: I would simply delete the text inside the tags and replace it with what I want the agent to say. Then click Save changes.

Example of what it would look like after:

<Response id="WhatsTheRate">
Your interest rate will be a reflection of your current credit standing as well as the year, make,
model, and mileage of the vehicle. We partner with 17 national lenders for exclusive rate breaks. I
just need to build your file and send it to the banks so we can get an accurate number for you.
That way, we are not misleading you in any way. Does that work for you?
</Response>

This editor is for more advanced editors who want to use AI to create and modify prompts. Proper prompt engineering techniques should be applied if using this. We recommend starting with the code editor to learn the prompt structure and best practices, then using the sidebar for guided changes.

Document view editor

The document view provides an easy-to-read view of the current prompt.

Steps to creating or Editing an AI Agent

Open /agents and choose Grid or List view.
Create New Process or edit an existing one.
Save your edits.
Test your agent in the playground.
Optionally publish a process to the Marketplace (superadmin only).
Confirm changes and verify agents behave as expected.

Common issues + fixes

Delete an agent blocked due to usage → use Reassign flow to move contacts.
Not seeing “Publish to Marketplace” option → requires Super Admin.
Agent not responding correctly → find the Agent in the “Agents” and head to the code editor. Hit (Ctrl/Cmd + F), find the objection handle and swap it with your desired output.
Agent using a different conversation flow → likely the wrong process (agent) is selected in the conversation view.

Escalation

Process bugs or degraded outcomes → share a screenshot of the response in the dealership’s WhatsApp group chat so Senior Prompt Editors can fix quickly.

Shortcuts/Links

Page: /agents
URL: https://app.automotiveai.ca/agents

Hotkeys in app

Open Command Bar: Cmd (Mac) / Ctrl (Windows) + K

Quick Reference

Page: /agents
URL: https://app.automotiveai.ca/agents

Keyboard Shortcuts

Open Command Bar: Cmd (Mac) / Ctrl (Windows) + K
Search in Editor: Cmd/Ctrl + F

Agents are powerful tools for automating and optimizing your customer interactions. Start with Getting Started to create your first agent, or explore the other sections above to dive deeper into specific topics.