Brisbane | AI Engineer Live Coding Session

Starting with project setup, watch an AI agent come together live—with information retrieval, tool calling, human approval and execution tracing.
Join us at JR Academy’s Brisbane office on 24 September to watch Lightman code in person and understand how an agent goes from code to a system that can carry out tasks.

## 🤖 From a Chat Demo to an Agent That Can Execute Tasks

Calling a model API, writing prompts and connecting a chat interface are the starting points of AI application development. When an agent begins handling real tasks, a series of engineering questions follows:

• Can its answers be backed by source material?
• Which tools should the agent have access to?
• How do you limit tool permissions and the number of calls?
• How do you stop or resume a task after an error?
• Which actions require human approval first?
• How do you record execution, evaluate results and investigate problems?

Through live coding, Lightman will address these questions within a complete agent project.

## 🚀 What You’ll See Live

### RAG & Citations | Make Answers Verifiable

Connect knowledge retrieval so the agent can answer questions using source material and provide references that can be checked.

### MCP Tools & Tool Calling | Connect External Capabilities

Give the agent access to tools and see how it selects a tool, passes arguments, reads the results and continues working on a task.

### Agent Loops & Execution Boundaries | Manage Task Execution

Break down the agent’s execution loop, set stopping conditions and execution limits, and handle incorrect tool calls and repeated execution.

### Human Approval | Keep People in Control of Important Steps

Add a review step before important actions and see how the agent pauses, waits for approval and then resumes execution.

### Failure Handling & Execution Tracing | Understand What Happened at Every Step

Inspect execution logs, analyse failures and learn how to locate problems, laying the groundwork for evaluation and post-run reviews.
Follow along as Lightman writes the code, connects tools and works through problems. Ask questions and discuss them with him in person.

## 🔥 Introducing the Upgraded AI Engineer Program — Cohort 07

This open session will also introduce the upgraded AI Engineer program: 13 weeks spent building and improving the same agent project, with each stage of learning contributing to one integrated system.
The project will progress through five system upgrades:

### ① System Foundation | Establish the Foundations

Define requirements, architecture, data and business workflows, and clarify the problem the project needs to solve.

### ② First AI Workflow | Get Your First AI Workflow Running

Integrate AI into a real workflow and produce structured outputs that can be checked and reviewed.

### ③ Grounded Intelligence | Build Evidence-Based Answers

Add RAG, citations, no-answer behaviour and quality evaluations so the system can reference source material and decline to answer when there is insufficient evidence.

### ④ Agent System | Build Complete Agent Capabilities

Connect MCP tools and strengthen execution boundaries, human approval and safe memory handling.

### ⑤ Production Standard | Add the Capabilities Needed for Production

Introduce harnesses, model routing, evaluations, red teaming and rollback to strengthen execution controls, assessment and release processes.
The final project will include code, architecture, evaluations, execution logs and release evidence, forming a complete agent system.

## 🧑‍💻 How We Teach: Understand the Principles, Then Apply Them to Your Project

The program includes two types of live sessions each week:

### Theory Live | Understand the Principles

Learn the core concepts, system architecture and engineering decisions involved in development.

### Practice Live | Build the Project

Through live coding, integrate each week’s capabilities into the project.
12 Theory Live sessions + 13 Practice Live sessions, continuously upgrading the same agent project from Week 1 to Week 13.

## 👥 Who Should Attend?

• Software engineers
• Backend and full-stack developers
• Data and ML engineers
• DevOps and cloud engineers
• Anyone who has built a RAG or agent demo and wants to develop their AI engineering skills further
• Technical professionals preparing to transition into AI engineering or interview for AI Engineer roles
• IT students with programming experience who want to understand practical AI engineering

## 💬 Meet Lightman in Person in Brisbane

Alongside the live development demonstration, we’ll set aside time for questions and conversation. Bring your technical questions, project ideas and learning plans, and discuss how to build on your current skills to develop the capabilities an AI Engineer needs.
You can also learn about Cohort 07’s project structure, learning schedule and prerequisites to decide whether the program fits your goals.

## 👨‍💻 Speaker

Lightman Wang | Founder of JR Academy
Lightman will write, run and explain the code live, walking you through the key steps and engineering decisions involved in agent development.

## 🗓️ Event Details

• Date: Thursday, 24 September 2026
• Time: 5:30–8:30 pm Brisbane local time (AEST)
• Venue: JR Academy Brisbane Office
• Address: L10b, 144 Edward Street, Brisbane City QLD 4000
• Speaker: Lightman Wang
• Format: In-person open session with face-to-face Q&A
• Admission: Free

## 🎟️ See How an Agent That Can Execute Tasks Is Built

Register below. See you in Brisbane on 24 September!