CTO rule number one: you are nothing without your team. I have believed that for a long time, but every so often a week comes along that proves it beyond argument. SideHack 2024 was one of those weeks.
We have just spent five days together in Portugal at our annual hackathon. Rob Harvey and I came home convinced we have the best team in the world. This post is an attempt to explain why, and to record what happened before the details blur.
What five days looks like
SideHack is a full working week, not an evening with pizza and a leaderboard. The whole engineering and product organization gets on planes, forms teams, and builds things that were not on any roadmap. The rules are deliberately light: pick a problem you care about, find people who care about it too, and have something real to show by the final afternoon.
Sidetraders in Portugal for the week, from every engineering and product team.
Projects delivered and demonstrated live on the final day.
Days, from first whiteboard to last presentation.
AI everywhere
Every year the projects tell you what the team is thinking about, and this year the answer was unambiguous. AI was the default rather than a track or a theme. Almost every one of the 18 projects used a model somewhere, and a good number of them used models to do work that a year ago we would have assumed needed a person.
The eighteen fell into three rough buckets, and the lines between them were blurrier than that makes it sound. Some were product prototypes, pushing the Order-to-Cash platform in directions we had talked about but never built. Some were engineering tooling, teams building for themselves: faster feedback loops, smarter automation, less friction shipping. That is the unglamorous work that compounds. The rest were built to be demonstrated live to the whole company, because slides-only entries are not allowed. The bar for a hackathon demo at Sidetrade is now uncomfortably high.
The ideas came from the people closest to the code and the customers, not from a strategy document, which is exactly how it should work.
The format is the point
People sometimes ask whether a week away from the roadmap is a cost we can afford. I think the question is backwards. The week pays for itself several times over, and not only in the projects that ship.
Nobody is assigned. People who have never worked together find each other around a problem, and those relationships survive long after the week ends.
Five days, a live demo, and an audience of 120 colleagues. Constraints force decisions, and decisions are where the learning is.
Half the value of a hackathon is discovering which ideas do not survive contact with a prototype. That is cheap to learn in a week and expensive to learn in a quarter.
When the demos land, everyone in the room understands what engineering is capable of. That changes conversations for the rest of the year.
Running, beaches and terrible karaoke
A hackathon that is only a hackathon misses most of the point. Being in the same place for five days meant morning runs and bike rides before the laptops opened, beach time in the afternoons, a Hawaiian night that will not be forgotten by anyone who has seen the photos, and a lot of late night dancing.
There was also karaoke. Some of it was terrible. All of it was worth it. Teams that have sung badly together in public tend to ship well together afterwards.
Hands down the best Hack we have run to date.
What I take home
The projects are what the week produced. What I take home is that 120 people picked their own problems, found each other without being assigned, and had something working by the final afternoon. I would not know how to buy that.
I am excited for the team, because they built it, and they know it. If you read this and wanted to be in the room, we are hiring.
