Join us on March 20th for a hands-on workshop introducing the core engineering patterns behind modern AI agents
Large language models have dramatically improved AI’s ability to generate text and code, but the most important developments now lie in turning models into systems. Modern AI agents extend language models with tool use, memory, and reasoning loops, allowing them to interact with APIs, operate software, and execute multi-step tasks autonomously. This shift—from models that generate answers to systems that take actions—is rapidly reshaping how AI systems are being engineered.
This workshop provides a hands-on introduction to the core engineering patterns behind AI agents. Participants will build a working agent and explore the architectural components that enable tool use, reasoning loops, persistent memory, and safety guardrails. By the end of the session, attendees will have a concrete understanding of how modern agent systems are constructed and why this emerging paradigm is quickly becoming central to both AI research and software engineering practice.
This event is supported by the Val O'Donovan Chair in Trustworthy AI.
Meet the workshop lead:
Hamza Mostafa is a 4th-year student in Computer Science at the University of Waterloo. He has interned at companies and startups in the San Francisco Bay Area, including OpenAI, where he worked on the Agent team building Deep Research, ChatGPT Agent, and Shopping Agent. Previously, he ran an AI consulting business helping enterprises leverage their data to build specialized agents.
Meet the host:
This event is hosted by Dr. Amir-Hossein Karimi, an Assistant Professor in the Department of Electrical and Computer Engineering at the University of Waterloo, and a Faculty Affiliate at the Vector Institute. Dr. Karimi holds the Val O'Donovan Chair in Trustworthy AI.
Workshop Roadmap:
| Part 0 | Part 1 | Part 2 | Part 3 | Part 4 | Part 5 |
|---|---|---|---|---|---|
| Setup | What are Agents? | Building Blocks | Build your Agent | Trust and Safety | What's next? |
| hands-on | teach | teach + build | hands-on | experiential | teach |
3 hours total | ~100 min hands-on
What You’ll Gain From This Workshop:
- A working AI agent you built yourself
-
Understanding of tool calling, memory, and the agent loop
-
Hands-on experience breaking and defending agents
-
A repo you keep building on after today
What You’ll Need to Participate:
- A laptop with WiFi
- Python 3.9+ (Parts 1-2)
- Node.js 18+ (Part 3)
- Telegram on your phone (Part 3)
- OpenAI API key: provided — you pay $0
When and Where:
Date: March 20, 2026
Time: 1:00pm to 4:00pm
Location: EIT 3142
Registration required!