GenAI Montréal @ Botpress
When AI Agents Fail — And What Actually Makes Them Work
GenAI Montréal is back with a special edition hosted at Botpress HQ, in partnership with Botpress and McGill AI Society.
This evening focuses on a simple but uncomfortable truth:
Most AI agents fail in production.
Not because the models aren’t powerful.
But because the system around them isn’t built to make them reliable.
We’ll explore what actually works when moving from demos to real-world autonomous systems.
***
## Evening Schedule — March 17
Botpress — 6:00 PM Doors Open
### 6:00 PM — Arrival & Networking
Light food & drinks
Meet members from GenAI Montréal, Botpress, and McGill AI Society
### ️ 6:30 PM — Welcome
- Introduction to GenAI Montréal
- A word from Botpress
- McGill AI Society introduction
- Overview of the evening
### 6:45 PM — Talks Begin
***
## Speaker #1
Michael Masson
CTO, Botpress
### Bio
As the CTO of Botpress, I build the infrastructure layer for autonomous AI agents, driven by an obsession of leverage and scalable architecture. Rooted in platform engineering, my philosophy is simple: build robust primitives that automate the routine so developers can solve the novel. We are currently shifting from softwares to products, and I'm dedicated to leading that transition. I'm an active advocate for the engineering community and a firm believer that the best ideas are born in the open. I'm always looking to connect with builders who are as excited as I am about the shift toward product oriented software.
### Talk
We Built the Engine That Makes AI Agents Actually Work
Abstract
We didn’t set out to rethink how AI agents work.
We just kept watching them fail in the same place, over and over.
This talk is the story of what was discovered, what was rebuilt, and why it changes what’s actually possible with autonomous AI systems.
Instead of chasing smarter models, this session explores:
- Where agents systematically break in real-world usage
- The architectural patterns that prevent repeated failure
- What it really takes to move from agent demos to production systems
- Why infrastructure and orchestration matter more than most teams realize
A behind-the-scenes look at building agent systems that don’t just impress—but operate reliably.
***
## Speaker #2
Morgan Willis
Principal Cloud technologist, AWS
### Bio
Principal Cloud Technologist @ AWS
Morgan Willis is a cloud and software engineering specialist with over 15 years of experience in tech, and more than 8 years focused on helping developers adopt cloud technologies through technical education.
Their background spans application architecture, backend development, and cost and performance optimization, with recent work centered on integrating generative AI and AI agent systems into production-grade environments.
Morgan creates technical content that bridges real-world engineering practices and developer learning, including courses on Coursera, tutorials, and live streams. Their focus is on showing how cloud and AI technologies can be applied securely, efficiently, and sustainably at scale.
### Talk
I Don't Trust AI Agents (And Neither Should You): Building Production-Ready Architectures
Abstract
Your AI agent works great in the demo. Then you deploy it and it hallucinates a refund policy that costs you $10K, or exposes customer data, or just loops endlessly burning tokens.
This session explores how to build AI agents you can trust in production using Amazon Bedrock AgentCore and the Strands Agents SDK.
We’ll walk through a layered approach to agent safety that spans guardrails at multiple levels, including input validation, output filtering, and action approval, along with observability patterns that make agent behavior transparent and auditable.
You’ll also see how multi-agent architectures can be designed so agents steer and check one another, and how security policies combined with deliberate prompt design reduce risk across the system.
We'll walk through a reference architecture that combines these layers, and you'll learn how each layer can catch failures.
You should have basic familiarity with LLMs and AI agents. You'll leave with architectures pattern you can adapt and a clearer picture of what can go wrong and how to address common issues with AI agent trust and reliability.
***
## Who Should Attend?
- AI engineers building agent systems
- Developers working with LLM frameworks
- Product & tech leads moving AI to production
- McGill students interested in real-world AI engineering
- Anyone curious about what separates working AI systems from hype
***
## Partners
Botpress — One of Montréal’s leading AI agent platforms
McGill AI Society — Connecting students with real-world AI innovation
Botpress works closely with McGill talent, including many interns from the university—making this meetup a natural bridge between industry and academia.
***
## About GenAI Montréal
GenAI Montréal is a 1,200+ member community organizing monthly in-person meetups focused on deep-tech, production-grade generative AI—from agent orchestration to agent infrastructure and beyond.
Join us for an evening grounded in reality:
What breaks. What scales. And what actually works.
