Introduction
Learn how to build a customer support AI agent with automated FAQ resolution and ticket escalation using Axellero Studio.
HelpDeskAI
HelpdeskAI is a customer support AI agent built in Axellero Studio that demonstrates how to automate first-line support using FAQs and seamlessly escalate unresolved issues into a structured ticketing system.
This tutorial is designed as a complete, end-to-end demo project. By following it step by step, you will reproduce the exact behavior shown in the recorded demo and understand how AI agents, tools, workflows, and databases work together inside Axellero Studio.
What You’ll Build
By the end of this tutorial, you will have built:
- An AI agent that understands user support requests
- Automated FAQ search and response logic
- Conditional escalation to ticket creation
- Customer identification and reuse
- Structured conversation logging for support teams
- A database-backed helpdesk workflow
1. Backbase
Persistent storage for customers, tickets, messages, and FAQs
2. Tools
Explicit backend actions such as search, create, and retrieve
3. AI Agent
Understands user intent and orchestrates workflows
4. HelpDeskAI
Combining tools and AI Agent
Demo
Demonstration of agent
Prerequisites
Before starting, make sure you have:
- Access to Axellero Studio
- Permission to create:
- Agents
- Tools
- Backbase database tables
- Basic understanding of:
- REST-style tools
- Databases and tables
- AI agent system prompts
This tutorial is beginner-friendly and does not require prior AI or ML experience.
High-Level Architecture
HelpdeskAI follows a tool-driven AI architecture, where the agent makes decisions but never performs system actions directly.
+-------------------+
| User |
+---------+---------+
|
+-------------------+
| HelpdeskAI |
| Agent |
+---------+---------+
|
+-------------------+
| Tools |
| (Search, Create, |
| Retrieve) |
+---------+---------+
|
+-------------------+
| Database |
| (Customers, |
| Tickets, FAQs) |
+-------------------+