University of Macau
University of Macau ยท Summer 2026

AI for the Real World
Code, Agents & Your Career

How the latest AI models are transforming software engineering, business operations, and the way we work โ€” featuring QoderWorks

3h
Duration
4
Modules
2
Live Demos
โˆž
Possibilities
Overview

Today's Journey

๐ŸŒ

Module 1

The AI Revolution

Why AI matters now โ€” from hype to production

45 min
โšก

Module 2

AI Software Engineering

QoderWorks in action โ€” live coding demos

50 min
๐Ÿค–

Module 3

AI Agents in Business

Beyond coding โ€” agents that transform workflows

40 min
๐Ÿš€

Module 4

Your AI Future

Careers, GBA opportunities & next steps

30 min
โ˜• 15 min break between Module 2 & 3
M1 ยท The AI Revolution
01
๐ŸŒ
Module 1 ยท 45 Minutes

The AI Revolution
in Business

Why AI matters NOW โ€” from hype to production

M1 ยท The AI Revolution
The AI Landscape 2025โ€“2026

From Chatbots to Agents

๐Ÿ• 2023 โ€” Generative AI

  • ๐Ÿ’ฌ Chat-based interactions only
  • ๐Ÿ“ Text & image generation
  • ๐Ÿ” Single-turn Q&A
  • ๐Ÿ‘ค Human drives every step
  • ๐Ÿ“‹ Copy-paste workflow

๐Ÿš€ 2025โ€“2026 โ€” Agentic AI

  • ๐Ÿง  Autonomous multi-step reasoning
  • ๐Ÿ› ๏ธ Tool use: code, browser, APIs, DBs
  • โ™ป๏ธ Self-correcting & iterative
  • โšก AI executes, human supervises
  • ๐Ÿ”„ End-to-end task completion
M1 ยท The AI Revolution
Where the Money Goes

Top 5 Business Functions Transformed by AI

๐Ÿ’ป

Software Dev

AI writes, reviews & deploys code 10ร— faster

๐Ÿ“Š

Data & BI

Natural language queries on real business data

๐ŸŽง

Customer Service

AI agents handle tickets end-to-end

โœ๏ธ

Content & Marketing

Campaigns, copy & media at scale

โš™๏ธ

Operations

Workflow automation across departments

0
% enterprises using AI
0
% productivity gain avg
0
$B global AI market (2026)
M1 ยท The AI Revolution
China's AI Ecosystem

China's AI Powerhouses

๐Ÿ”ฎ

DeepSeek

Open-source LLMs rivaling GPT-4 at a fraction of the cost. Disrupted the industry with efficient training.

Open Source Reasoning
๐Ÿง 

Alibaba Qwen

Full-stack AI: language, vision, audio. Powering Alibaba Cloud enterprise services globally.

Enterprise Multi-modal
๐Ÿค–

Baidu ERNIE

Knowledge-enhanced models integrated into Search, autonomous driving & smart devices.

Knowledge Industry

๐Ÿ’ก Key Insight: China is not just catching up โ€” it's leading in open-source models, AI application layers, and cost-efficient training. The global AI race is now a collaborative competition.

M2 ยท AI Software Engineering
02
โšก
Module 2 ยท 50 Minutes

AI-Powered
Software Engineering

Inside QoderWorks โ€” how AI is transforming the way software is built

M2 ยท AI Software Engineering
Why QoderWorks

The Best AI Agent for Business

QoderWorks isn't just another coding tool โ€” it's a full-stack AI agent platform built for real-world business impact.

๐Ÿง 

Agentic, Not Just Generative

Unlike ChatGPT or Copilot, QoderWorks doesn't just generate text โ€” it autonomously plans, executes, and verifies multi-step tasks end-to-end. It thinks, acts, and self-corrects.

๐Ÿ”Œ

MCP-Native Architecture

Built from day one on the Model Context Protocol โ€” connect AI to any database, API, or SaaS tool instantly. No custom integrations needed. The "USB-C" of AI is already built in.

๐Ÿข

Enterprise-Ready from Day One

Production-grade security, audit trails, role-based access. Not a toy or research prototype โ€” deployed in real businesses handling real data and real workflows today.

โšก

10ร— Faster Than Traditional Dev

From idea to deployed app in minutes, not days. QoderWorks handles architecture, coding, testing, and deployment โ€” a full engineering team in one agent.

๐ŸŒ

Beyond Code โ€” Full Business Agent

Browser agents for research, data agents for BI, document agents for legal. One platform, every knowledge workflow โ€” not limited to just writing code.

๐Ÿค

Built by Exeurion for Real Business

Developed with deep understanding of enterprise coaching, sales, and retail operations. Not built in a lab โ€” built by solving real business problems for real clients.

M2 ยท AI Software Engineering
The Problem

The Software Engineering Crisis

Businesses are desperate for speed. AI coding assistants are the answer.

1.4M
Unfilled developer jobs globally
85%
Of projects delivered late
$300B
Lost to tech debt yearly
M2 ยท AI Software Engineering
Business Scenario

The Problem: Running a Clothing Store

๐Ÿ‘• The Business

A clothing retail store with 3 locations selling fashion to young professionals. Business is growing, but operations are a mess.

๐Ÿ˜ค The Pain Points

  • Inventory tracked in spreadsheets โ€” staff don't know what's in stock until they check the back room
  • No way to see which products sell best across stores โ€” bestsellers run out, dead stock piles up
  • Building a custom inventory system would cost $30K+ and 3 months with a dev team

๐ŸŽฏ Business Objectives

  • Dashboard showing real-time inventory across all 3 stores
  • Low-stock alerts so popular items get reordered in time
  • Sales tracking: what sold, where, and when โ€” at a glance
  • Weekly sales reports generated automatically for the owner

๐Ÿ’ก The ask: Build this app in minutes, not months โ€” let's see AI do it live.

M2 ยท AI Software Engineering
Live Demo 1
LIVE

From Business Need to Working App

๐ŸŽฏ What We'll Build

A full-stack Retail Inventory & Sales Dashboard โ€” built entirely from a natural language description.

Step 1

Describe the app

Tell Qoder the business requirements in plain English

Step 2

AI plans the architecture

Dashboard UI, product database, inventory tracking, sales reports

Step 3

Code generation

React frontend, Node API, SQLite database โ€” all auto-generated

Step 4

Live preview

Working inventory system in the browser โ€” in ~3 minutes

๐Ÿ”‘ Watch For

  • How AI plans before writing a single line of code
  • How AI creates files and structures a full project
  • How AI self-corrects when it encounters errors
  • Speed: A $50K app built in ~3 minutes

๐Ÿ’ก Traditional timeline: 3 months, 2 developers, $30K+

M2 ยท AI Software Engineering
Under the Hood

How an AI Coding Agent Works

๐Ÿ’ฌ
Prompt
User describes task
โ†’
๐Ÿง 
Plan
AI breaks down steps
โ†’
๐Ÿ”
Search
Reads codebase context
โ†’
๐Ÿ’ป
Code
Writes & edits files
โ†’
โœ…
Verify
Tests & self-corrects

๐Ÿง  Context Engine

Reads entire codebases, understands file relationships, dependency graphs, and project conventions.

๐Ÿ› ๏ธ Tool System

Executes terminal commands, edits files, browses the web, queries databases โ€” all autonomously.

๐Ÿ”„ Self-Correction

Detects errors, analyzes failures, tries alternative approaches โ€” without human intervention.

M2 ยท AI Software Engineering
Real Business Impact

Case Studies

๐Ÿš€

Startup MVP

Challenge: Build an MVP in 2 weeks with 3 people.

Result: AI-assisted dev shipped MVP in 4 days. Raised seed round in a month.

5ร— faster delivery
๐Ÿข

Enterprise Modernization

Challenge: Migrate 500K+ lines of legacy Java.

Result: AI analyzed, refactored & tested. 70% reduction in migration effort.

$2M saved
๐Ÿ”„

AI-Powered DevOps

Challenge: Reduce deployment failures & incident time.

Result: AI auto-fixes CI/CD, resolves incidents. 90% fewer deployment failures.

99.9% uptime
M2 ยท AI Software Engineering
โ˜•

15 Minute Break

Stretch, grab some coffee, and think about what you'd build with AI

95
minutes completed
โ†’
 
85
minutes remaining
M3 ยท AI Agents in Business
03
๐Ÿค–
Module 3 ยท 40 Minutes

AI Agents
in Business Operations

Beyond code โ€” agents that transform every knowledge worker's job

M3 ยท AI Agents in Business
The Rise of AI Agents

Chatbot โ†’ Copilot โ†’ Agent

๐Ÿ’ฌ

Chatbot

Single-turn responses. Human drives everything. Think: early ChatGPT.

Passive No memory
๐Ÿค

Copilot

AI assists in real-time. Human and AI collaborate side by side.

Collaborative Context-aware
๐Ÿค–

Agent

Autonomous execution. AI plans, uses tools, self-corrects. Human supervises.

Autonomous Tool-using

"The future isn't AI replacing humans โ€” it's AI agents working alongside humans, handling complexity at superhuman speed."

M3 ยท AI Agents in Business
Real-World Agents

AI Agents Across Industries

๐ŸŒ

Browser Agent

Automates web research: browses sites, extracts data, fills forms, compares products. Handles complex multi-step web tasks autonomously.

Market Research Competitor Analysis
๐Ÿ“Š

Data Agent

Connects to databases, writes SQL, generates charts & reports. Turn "show me Q2 sales trends" into instant visualizations.

Business Intelligence Reporting
๐ŸŽง

Customer Service Agent

Handles support tickets end-to-end: reads context, diagnoses issues, resolves problems, and follows up โ€” all without human handoff.

Support Resolution
๐Ÿ“„

Document Agent

Drafts contracts, reviews legal docs, extracts key clauses, flags risks. What took lawyers hours now takes minutes.

Legal Compliance
M3 ยท AI Agents in Business
Business Scenario

The Problem: Staying Competitive in Phone Retail

๐Ÿ“ฑ The Business

A mobile phone retail store wants to know what competitors are charging and which phone models are trending โ€” so they can price smart and stock what customers will want next.

๐Ÿ˜ค The Pain Points

  • Staff spend hours every week manually checking competitor websites for prices and promotions
  • Data comes from 10+ sources โ€” JD.com, Taobao, competitor stores, tech review sites, social media
  • By the time a pricing report is done, competitors have already changed their prices
  • Hiring a market research firm costs $5Kโ€“$15K per report

๐ŸŽฏ Why an AI Agent?

  • Autonomous web browsing โ€” the agent visits competitor sites, extracts phone prices, specs, and reviews on its own
  • Multi-source comparison โ€” combines JD.com, Taobao, and store data into one price comparison table
  • Trend detection โ€” identifies which phone models are gaining buzz and which are declining
  • Repeatable โ€” run the same agent daily for always-fresh competitive intelligence

๐Ÿค– Let's build this agent live โ€” and watch it work in ~2 minutes.

M3 ยท AI Agents in Business
Live Demo 2
LIVE

Building a Market Research Agent

๐ŸŽฏ The Mission

Build a Market Research Agent that autonomously:

Step 1

Search the web

Find competitor information online

Step 2

Extract data

Pull key data points from multiple sources

Step 3

Analyze trends

Identify patterns and competitive insights

Step 4

Generate report

Produce a structured, actionable report

๐Ÿ” Agent's Thinking Process

Agent Internal Reasoning
// Agent's internal reasoning:
Plan: Research "AI code editors" market

  โ†’ search_web("top AI code editors 2026")
  โ†’ browse("competitor-site.com/pricing")
  โ†’ extract(pricing, features, reviews)
  โ†’ analyze(comparison_data)
  โ†’ generate_report(structured_output)

// Confidence: HIGH
// Sources checked: 12
// Total time: ~2 minutes โšก

Output: {
  competitors: 8,
  insights: 15,
  recommendation: "..."
}
M3 ยท AI Agents in Business
AI + Business Data

MCP โ€” The Universal AI Connector

Model Context Protocol: How AI connects to everything

๐Ÿ—„๏ธ
Database
โ†’
๐Ÿ”Œ
MCP Server
โ†’
๐Ÿง 
AI Agent
โ†’
๐Ÿ“Š
Insights

๐Ÿ—„๏ธ Data Sources

Databases, APIs, file systems, SaaS tools โ€” all connected through standardized MCP servers.

๐Ÿ” Security

Enterprise-grade access control. AI only sees what it's authorized to. Full audit trail.

๐ŸŒ Open Standard

MCP is open source โ€” any AI model, any tool can plug in. The "USB-C" of AI integration.

M4 ยท Your Future with AI
04
๐Ÿš€
Module 4 ยท 30 Minutes

Your Future
with AI

Career paths, skills, and the road ahead

M4 ยท Your Future with AI
The AI-Augmented Professional

New Roles & Skills

๐Ÿ†• Emerging Roles

  • AI Engineer โ€” Build & deploy AI-powered apps
  • Prompt Engineer โ€” Design optimal AI interactions
  • AI Product Manager โ€” Define AI-driven products
  • AI Ops โ€” Monitor & optimize AI in production
  • AI Ethics Lead โ€” Ensure responsible AI use

๐Ÿ’ก Skills That Matter Now

  • Problem decomposition โ€” Break complexity into AI-sized tasks
  • System thinking โ€” Design AI-human workflows
  • AI collaboration โ€” Know when to trust, when to override
  • Domain expertise โ€” AI amplifies experts, not generalists
  • Critical evaluation โ€” Verify AI output quality

"AI won't replace you โ€” but someone using AI will."

M4 ยท Your Future with AI
GBA Opportunities

AI in the Greater Bay Area

๐Ÿ›๏ธ

Macau

Tourism AI, cross-border fintech, academic research hub

๐ŸŒ†

Zhuhai

AI manufacturing, smart city labs, Hengqin innovation zone

๐Ÿ™๏ธ

Shenzhen

Hardware + AI, robotics, Tencent HQ, startup capital

๐ŸŒƒ

Hong Kong

AI in finance, legal tech, international R&D hub

๐Ÿ’ฐ Government Funding

GBA AI innovation funds, startup grants, research subsidies across all four cities.

๐Ÿค Cross-Border Collaboration

One region, two systems, multiple currencies โ€” ideal for AI fintech experiments.

๐ŸŽฏ Hot Industries

Fintech, tourism tech, smart city, healthcare AI, education tech, logistics.

University of Macau

The Future is
Agentic AI

Stay curious. Stay ahead. Here's where to keep learning.

โšก

Explore QoderWorks

Try the AI agent platform featured in today's workshop. Build apps, automate workflows, and see agentic AI in action.

qoder.ai Free to try
๐Ÿ“ฐ

AI News in China

Stay up to date with the latest AI breakthroughs, industry news, and research from China and beyond.

ๆœบๅ™จไน‹ๅฟƒ jiqizhixin.com ้‡ๅญไฝ qbitai.com
๐Ÿง 

Open Source AI Models

Explore China's leading open-source LLMs โ€” try them, fine-tune them, and build your own AI-powered projects.

DeepSeek deepseek.com Qwen qwen.ai
๐ŸŽ“

Learn & Build with AI

Hands-on tutorials, communities, and tools to start building AI apps today โ€” no PhD required.

Hugging Face huggingface.co ModelScope modelscope.cn

Thank you! ่ฐข่ฐข!

โ† โ†’ or Space to navigate ยท F fullscreen