Agentic AI: The Orchestration Shift

## 🤖 GenAI Montréal @ ConFoo

### Agentic AI in Production: The orchestration shift

GenAI Montréal is hosting a special ConFoo edition meetup in a multi-community breakout format, bringing together Montréal’s Java, Cloud-Native, AI, and CTO communities for one shared evening—then diving into focused tracks.

This session is about what actually works in production when building agentic systems: architecture, routing, and knowing the limits of your AI.

***

## 🕔 Evening Schedule (Special ConFoo Format)

### 🍕 5:00 PM — Welcome, Beer & Pizza

📍 Saint-Laurent 1
Casual networking with attendees from all participating communities.

***

### 🎙️ 5:30 PM — Opening & Introductions

📍 Saint-Laurent 1

• Welcome to the evening
• Overview of the multi-community format
• Introduction of group leaders
• Acknowledgment of sponsors: AWS & MySQL

***

### 🔀 5:45 PM — Community Breakouts Begin

Attendees split into parallel sessions by interest:

| Community | Room |
| --------- | ---- |
| MTL JUG | Saint-Laurent 1 |
| CNCF Montréal | Saint-Laurent 4 |
| GenAI Montréal | Saint-Laurent 5 |
| MTL CTO | Saint-Laurent 6 |

***

## 🧠 GenAI Montréal Track

📍 Saint-Laurent 5

### 🎤 Speaker #1

Nicholas Nadeau
Cofounder, Onix · Ph.D. Robotics & AI
Nicholas Nadeau works on human-system collaboration—how people and intelligent systems work together effectively.

• Former CTO at 1X (humanoid robots deployed alongside people)
• Former CTO at AON3D (industrial 3D printers used for ISS & Moon missions)
• Former CTO at **SmartOne.ai** (human-in-the-loop enterprise AI)
• Cofounder of Onix, building Personal Intelligence: AI that works with you, not on you
• Cofounder of AI Salon Montréal
• President of the Board at Garage&Co.
• Mentor at Creative Destruction Lab and Next AI Montréal

📍 Based in Montréal

***

### 🧩 Talk

Agentic AI Is a Routing Problem, Not an Intelligence Problem
Abstract
Everyone is racing to make agents smarter—bigger models, more tools, more reasoning.
But agents that survive production aren’t the smartest ones.
They’re the ones with the best routing.
Agents don’t fail because they can’t reason.
They fail because they don’t know what they don’t know.
This talk focuses on the architectural patterns that actually work:

• Routing as the core primitive of agentic systems
• Expertise graphs instead of brittle chains of thought
• Embeddings for intelligent dispatch
• The “know your limits” pattern for agents that recognize the boundaries of their competence

A practical, production-oriented perspective on building resilient agentic systems.

***

### 🎤 Speaker #2

Andres Ortega
Data Scientist → Full-Stack Developer → Kubernetes Engineer
Andres has worked across the full AI stack, from data science to platform engineering. He has contributed to large-scale AI and ML foundations powering major social media platforms, with a strong focus on production infrastructure and security.

#### 🧩 Talk

Securing AI Workloads — Present and Future
Abstract
AI-generated workloads are accelerating at an unprecedented pace.
Code reviews are becoming less effective.
Supply-chain attacks are more common—and more subtle.
In this talk, we explore modern security challenges unique to AI systems, including:

• Preventing data exfiltration in AI workloads
• Defending against prompt injection attacks
• Securing AI supply chains in a world of generated code and models

A forward-looking discussion on how to secure AI systems as they move from experimentation to critical production infrastructure.

***

## 👥 Who Should Attend?

• AI engineers & ML practitioners
• Software developers exploring agentic systems
• Architects & tech leads
• Researchers interested in human-AI collaboration
• Anyone building GenAI beyond demos

***

## 🌐 About GenAI Montréal

GenAI Montréal is a 1,100+ member community organizing monthly in-person meetups focused on deep-tech, real-world generative AI—from agentic systems to production architectures.