Students from our Computer Engineering Technology – Computing Science program recently placed 2nd at the DevOps for GenAI Hackathon, held as part of AWS Community Day Ottawa 2026. This is an event where “…builders, engineers, and technology leaders come together to explore the future of cloud and AI-powered delivery.”
Hear more about it directly from one of our participating students, Abdirahman Dahir:
“18 teams. Around 200 people in the room. Industry judges and mentors watching over our shoulders the whole time. Walking in, I was nervous — not just because it was my first hackathon, but because I was carrying the responsibility of leading a team.
Leaning on what I’ve learned from past team projects, from leadership across the last few semesters, and from my time as an IBM Ambassador helped me navigate the pressure. And the volunteers and mentors made every one of us feel welcome from the moment we arrived.
We competed as The Hallucinators, a name we loved, because “hallucination” is the exact failure we set out to hunt in AI.
The challenge was a red-team project: “Break My AI.” Under real-time pressure, we built a document-aware RAG chatbot running a Llama model locally through Ollama, and then came the fun part: we attacked it. Prompt injection to hijack the model’s instructions. Data-leakage attacks to make it reveal what it shouldn’t. RAG document-poisoning to smuggle malicious instructions inside uploaded files. And chaos attacks to break it with bad input. Every result rolled up into a live security score.
The hardest part wasn’t the code. It was putting ourselves out there and taking that first step, then delivering everything under the clock and presenting it in just 4 minutes. We pushed through that and a dozen smaller obstacles along the way.
Here’s what stuck with me: teams are shipping AI into the real world without ever attacking it first. We wanted to build the thing that breaks the AI before real users or attackers do.
Two days of learning you just don’t get in a classroom.
To anyone still waiting for the right moment to take their first step — consider this your sign.”
We Value Your Privacy
The Algonquin College website uses cookies to enhance your browsing experience. The cookies categorized as "Essential" are stored on your browser and enable the basic functionalities of the website. We also use third-party cookies to help us analyze how you use the website to maximize performance, store your preferences, and provide you with relevant content and advertisements. You can choose to allow all cookies or only allow a selection, but disabling some of them may affect your browsing experience. View our privacy statement.
These cookies cannot be turned off as they are required to enable the basic features of this site. These cookies do not store any personally identifiable data.
These cookies are used to collect preferences, compile site usage information, and improve the efficiency of the website.
These cookies allow us and partner companies to make our marketing efforts more relevant and to optimize the effectiveness of our ads.
Students from our Computer Engineering Technology – Computing Science program recently placed 2nd at the DevOps for GenAI Hackathon, held as part of AWS Community Day Ottawa 2026. This is an event where “…builders, engineers, and technology leaders come together to explore the future of cloud and AI-powered delivery.”
Hear more about it directly from one of our participating students, Abdirahman Dahir:
“18 teams. Around 200 people in the room. Industry judges and mentors watching over our shoulders the whole time. Walking in, I was nervous — not just because it was my first hackathon, but because I was carrying the responsibility of leading a team.
Leaning on what I’ve learned from past team projects, from leadership across the last few semesters, and from my time as an IBM Ambassador helped me navigate the pressure. And the volunteers and mentors made every one of us feel welcome from the moment we arrived.
We competed as The Hallucinators, a name we loved, because “hallucination” is the exact failure we set out to hunt in AI.
The challenge was a red-team project: “Break My AI.” Under real-time pressure, we built a document-aware RAG chatbot running a Llama model locally through Ollama, and then came the fun part: we attacked it. Prompt injection to hijack the model’s instructions. Data-leakage attacks to make it reveal what it shouldn’t. RAG document-poisoning to smuggle malicious instructions inside uploaded files. And chaos attacks to break it with bad input. Every result rolled up into a live security score.
The hardest part wasn’t the code. It was putting ourselves out there and taking that first step, then delivering everything under the clock and presenting it in just 4 minutes. We pushed through that and a dozen smaller obstacles along the way.
Here’s what stuck with me: teams are shipping AI into the real world without ever attacking it first. We wanted to build the thing that breaks the AI before real users or attackers do.
Two days of learning you just don’t get in a classroom.
To anyone still waiting for the right moment to take their first step — consider this your sign.”