— Experience
CxC
Waterloo’s largest AI hackathon, planning the next iteration
Highlights
- Co-chair Waterloo’s largest AI / data hackathon with 300+ students building on campus for a weekend
- Planning the next brief so teams can ship predictions, trained models, or data-science sites without a forced project type
- Planning workshops, company booths, and socials so people actually meet instead of shipping alone
- Owning the floor when it gets late: pairing builders, unblocking demos, and fixing whatever breaks before judging
How I got here
I did not join CxC to put “Co-Chair” on a LinkedIn line. I joined because Waterloo’s data and AI crowd is huge, and for a long time the biggest weekend on campus still felt like it was talking to a narrower slice of that crowd than it could.
CxC is Waterloo’s largest AI hackathon. Fifth year now. Hundreds of students, one campus weekend, mentors walking the floor, and a judging room that fills with demos that barely slept. I co-chair it. That means I care about the brief, the sponsors, the workshops, and whether a first-year who showed up nervous on Friday still wants to come back on Sunday.
What was broken
Older briefs leaned toward a specific kind of project. Fine if you already knew the genre. Rough if you had a weird idea, a half-trained model, or a dataset you actually cared about and did not want to force into last year’s shape.
People also showed up, built in a bubble, and left. The weekend worked as a shipping contest. It did not always work as a place to meet the person who would unblock you at 1am, or the company you might actually want to talk to after the closing ceremony.
What we’re planning
The brief
For the next iteration I want to open it. No project-type lock. Predictions on a dataset, a model you trained yourself, a small data-science site that does one job well: all of it should count. The point is to stop rewarding the team that mirrors the prompt the hardest and start rewarding the team that ships something real.
That decision sounds small on a landing page. On the floor it should change who feels invited. I want teams pitching ideas they would have self-rejected under a tighter brief, then actually building them.
The weekend around the build
A hackathon that is only laptops and silence is a study hall with worse chairs. The plan is to fill the gaps: workshops people can walk into without a perfect background, company booths that are not just swag tables, and socials that give introverts a second door into the room.
My job in that mix is less “announce from a stage” and more “make the room work.” Pair someone stuck on setup with someone who just finished it. Move a workshop when the room double-books. Notice when a booth is empty because the power strip died, not because nobody cared.
Late night
If you have never run a hackathon past midnight, it is hard to explain how physical the role gets. Cables. Room swaps. A demo laptop that will not join the AV. Someone whose teammate went home and who is quietly spiraling about presenting alone.
I stay on the floor for that part on purpose. Judging only works if the path to the demo still exists at 2am. Keeping the room loud in the right way (curious, not chaotic) is character work as much as logistics: stay calm, stay useful, do not make a tired builder feel stupid for asking the same question twice.
What I’m aiming for
Co-chairing CxC keeps teaching me that “AI event” is the easy label. The real product is whether people leave feeling like they belong in the room they just built in. Opening the brief is a product decision. Walking the floor at 2am is the delivery.
I am still in the role. The event keeps growing. The part I care about most has not changed: a weekend where Waterloo’s AI and data people get to build what they actually want, and still find each other while they do it.