University is My Side Hustle: How Gen Z Builders Ship to Production

Hey everyone, my name is Artemis. I'm a founding engineer at Canoparty and a second year university student at Computer Science in NKUA in Athens. Today I'm going to be talking to you about how universities are my side hustle. As you know now with AI, universities are kind of falling behind in this era because at our work we can use AI freely, but in universities we can't. It's still actually forbidden to use AI in your assignments and so on. So there are two completely different things. And I'm going to be talking to you today about some things, startup culture, Gen Z, how we blend into this AI era and we grow with it.

So basically, we will talk about how Gen Z builders are stepping into production while everyone else is waiting for permission. They're waiting for their degree. Of course, every option is valid, but let's get into it.

So we have two paths right now. The first one is the traditional path, which is what people used to do almost always, which is to go to the university, get a lot of theory, study a lot, get unpaid internships just for validation and just to have something on your resume. Then get some junior role, wait for someone to give you a zero ticket to work.

And basically, you have a lot of years of preparation, getting to corporate environments before you make a real impact and before you make something you actually like. But nowadays, especially this past year, we also have the builder path, which is something a lot of people do, especially in San Francisco. And it's basically learning via immediate deployment. You just write code with AI tools. You use whatever you have in your abilities to make something that you actually like, that people want, and deploy it online.

See how people like it. Get founding roles at a startup and basically have impact on production from day one. You can now use AI for the development, the marketing of your idea and everything. So it basically accelerated everything for us. So basically, what happened that allowed us to do this change is that the gatekeepers left, all the barriers are gone. You have AI at your fingertips. You can use a lot of modern tools for free.

So a 90-year-old student in a dorm room can now architect, deploy and scale systems that previously required an entire enterprise IT department. This has made things so easy for us, young people who want to make a change, who don't want the traditional path from junior to senior and so on. So we can now make our own startups, build our own ideas, see them live, try them out and basically experiment a lot. And this talk will be a lot about experimenting.

So the philosophy is that impact is over the grades. University isn't a priority anymore. It's kind of the playground. University is great, like Lucy Guo says, to find your future co-founders, to find your future employees, or the people who will hire you, et cetera. So it's a great opportunity to make friends, but not so much to learn how production works. You can't learn, especially in computer science, you can't learn how a user actually uses your product. But when you deploy something to production, you can see how users interact, what do they need, and how is what you build relevant to their needs and wants.

The real degree now is your production code base. It is your GitHub profile. It is not so much your assignments and your theoretical practice. The whole mindset is don't ask for permission to build the future. The deployment button doesn't check your degree. If you go to AWS and try to bring up a server, you're not asked for a degree from the university. You can just try things out, experiment, and build things that actually might matter to some people. Right now, we cannot, as young people in their parents' house or whatever, ship something better specifically than a big corporate enterprise. However, we have a mode which is very important, and that is speed. We can ship it broken and get one user's feedback. This is even more worth it than a thousand local tests. Production is the best debugger.

As builders, we seriously have this problem where we create a lot of things for ourselves, and we fix our own problems. We have them locally on our computer, and we tend to over-engineer them a lot. And then we don't ever ship them to production. We never market them so nobody else can use them, which is good for a lot of things, especially if you have some projects that are very particular to your needs. However, you can create something more out of it by creating a startup now that we have all these tools available.

Also, we have iteration velocity, which is another big advantage for us. The founder who ships every 48 hours will always outrun the founder who ships every quarter. We know how this is true. We know how Cloud Code can help us achieve this. I'm pretty sure most of you have experimented with this. Also, we have execution over code. You can leverage AI to handle the boilerplate so you can focus 100 percent on core product strategy. This is the hardest part, in my opinion. Marketing and selling the product might be the hardest part. But building it has never been more fun. It has never been easier.

So let's go on. Where the effort goes now. Like I said, only 15 percent goes on to implementation. 30 percent is strategy and vision, and 55 percent, the hardest part, is distribution and sales. If your startup fails today, it's not because of a bug. It's because of your strategy, most possibly. Because if you build a startup in 2026, even if you get a bug, you can tell Cloud Code to fix it. It might not fix it perfectly or it could remove the feature, but you don't care because you spend so little time and effort and money to build it that your mistakes you make are not actually that expensive. So if you focus on distribution, strategy, and the vision of where you want the company to go, who do you want to see this? Who do you want to use it? And so on. However, bottling that we, most of us have, especially at a young age, and especially we have English as our second language and we don't live in San Francisco, is the communication.

We have a clarity gap because as engineers, we tend to overcomplicate things and ramble. So we need to find simplicity. I was at the YC event once and there was a Q&A and the people who were asking the questions, literally, you could not understand the question. They had to talk for about, they had to say about three sentences to explain what they want to hear from the speaker. So, especially for us, who have English as their second language, we need to learn how to communicate clearly and be able to say what we do in one sentence only. So, tech gets you in the room, but conviction keeps you there. You have to sell the vision before the code is ever written. That's why a lot of people say sell and then build. You can, for example, sell a product to somebody, ask them what they like, make it just a mock-up on lovable and then ship it later with better code.

So, make sure that we don't, due to AI, lose our soft skills, because this is a trap that a lot of people go through. Because we have AI that can write our emails and can write our LinkedIn posts, we forget that we also need to communicate in the real world sometimes. And like I said on LinkedIn, LinkedIn has now become a distribution weapon. I don't know if you've seen this, but there are a lot of young founders who just build in public and just post everything on LinkedIn. Even zip posting on LinkedIn. And it's not just a digital resume anymore. It's not just somewhere where you list all of your experience. It is a high leverage distribution channel for builders. Every feature shipped is a networking opportunity. So, you can show the work, explain the struggle and you get both the other founders who can help you. You get the VCs that can see what you're doing and help you out financially. And you can also get engineers or somebody else who you might want to hire, who might need your help later in the future. You might find even your sale target.

So, in general, building in public is a very good idea if you don't want to do the traditional Instagram marketing and it depends basically on your product. So, we have a lot of tools that can help us build in a very high velocity. What I have listed here is the more low level ones for people that haven't really written a lot of code. And it could be Bolt, V0, Cloud and Vercel. Vercel is more for hosting and infra, but still, you have a zero conflict production environment. If you have no idea about AWS and cloud, you don't really need it anymore. You can use Vercel and Superbase. So, there's really no excuse.

Even Cloud Code, if you ask it to set it up for you all, it can do so. And there are a lot of people who are not technical at all and they're building products. So, I don't get why us who like building are not doing the same and making great products for whatever we want.

Another thing that is very related, of course, to today's summit, which is React Summit, is that React actually published its own React best practice for AI. So, this is our cheat code that we will use to create our product. We can get the React best practices online and we can get them directly into our Cloud Code environment or our cursor. So, we use these specific instructions to ensure that our AI system follows clean architecture, proper hooks usage, and performance patterns.

Prior to that, we had seen a lot of vulnerabilities while some people were vibe-coding React code. But now I think this will help a lot. So, we've moved to write the code, to architecting the results. You can also write your own instructions if you want, custom instructions about your system design and whatever else. Everything that I have discussed about today is not something that I just read in a blog post and decided to share out loud. It's something that we're doing kind of party every day.

We're bringing people together at kids parties and we're part of planning fiction with high velocity shipping. We're interacting with real users. We see how they use the product. We see what their favorite pages are, for example. And make it easier for them every day.

There's not a single day in Canoparty where we haven't had a vibe-coded, at least, feature shipped. So, the impact is successfully deployed and scaling in the real world. It's not only Canoparty that does this, but a lot of companies and startups as well. It involves young people balancing university and work, becoming great founding engineers.

If you want to pursue this path, being close to other founders and engineers who prioritize fast shipping is crucial. Surround yourself with people who are bold in their decisions, unafraid to break things in production, and not bound by perfectionism. The industry may expect you to be a junior, but your execution velocity and vision can prove them wrong. Embrace the challenge to ship one thing, regardless of its size, and build in public with audacity and boldness.