From LLM to Agent
Download the PowerPoint slides for this course. Anyone who could not attend the class can review the main ideas, examples, and AI agent concepts covered in the session.
Date: July 5, 2026 (Sunday) | Time: 6:30 PM - 8:00 PM | Location: 76 Larkin Dr, Nepean, ON
Event Poster
Course Slides
AI Key Concepts Introduction
LLM, Token, Context, Prompt, Tool, MCP, Agent
Build AI-Powered Business Agents with Codex, including live running agent demonstrations.
Large Language Models and AI Agents are rapidly transforming how businesses operate, how software is built, and how people interact with technology. This session connects the core ideas into one clear mental model of how today's AI systems actually work.
In addition to the theoretical concepts, the session will include live demonstrations of AI agents and show how modern coding agents such as Codex can be used to build AI-powered business solutions.
Event Details
- Date: July 5, 2026 (Sunday)
- Time: 6:30 PM - 8:00 PM
- Location: 76 Larkin Dr, Nepean, ON
- Speaker: Fenqiang Zhuang, Founder & CEO of FeyaSoft Inc.
Topics Covered
Large Language Models
What is an LLM, how responses are generated, and why LLMs differ from traditional software.
Tokens and Context
Tokenization, context windows, and why AI systems sometimes forget information.
Prompt Engineering
User prompts, system prompts, instruction design, and practical prompt techniques.
Tools and Function Calling
How AI accesses data, APIs, and business systems through external tools.
MCP
Model Context Protocol and secure connections between AI and enterprise systems.
Agents and Skills
Agent planning, reasoning, skills, capabilities, and multi-agent collaboration concepts.
Business Agents with Codex
Coding agents, tool workflows, MCP-enabled business agents, and automation.
Complete AI Architecture
Connecting LLM, prompt, context, tools, MCP, skills, agents, and enterprise apps.
LLM -> Prompt -> Context -> Tools -> MCP -> Skills -> Agents -> Enterprise AI Applications
Live Demonstrations
- AI business assistant
- Enterprise data retrieval agent
- Tool-calling agent
- MCP-enabled business agent
- AI coding assistant examples
- Agent workflows and reasoning demonstrations
- Business automation scenarios
- Codex-powered solution building
Who Should Attend?
This session is designed for software developers, IT professionals, business leaders and managers, entrepreneurs and startup founders, students interested in AI, and anyone curious about how ChatGPT and AI Agents work. No prior AI experience is required.
About the Speaker
Fenqiang Zhuang is the Founder and CEO of FeyaSoft Inc., a Canadian software company specializing in CRM solutions, low-code platforms, workflow automation, and AI-powered enterprise applications.
With extensive experience in enterprise software, AI integration, and business automation, he has led the development of AI-enabled business platforms that connect large language models with real-world enterprise systems.