My AI Hackathon experience

No LLM was used, harmed, or jailbroken while writing this article. With memes because they said that Hackathons should be fun.

What the article is about

EPAM Product Management practice held an AI Hackathon in the mid of July 2023. Around 20 teams participated, and yes, the Hackathon was about AI and all this fancy stuff you read every day in the news, starting with ‘ChatGPT this,’ ‘ChatGPT that,’ and so on.

In this article, I wanted to share my experience throughout the whole thing and give you an idea of why you should consider doing a Hackathon once in a while.

How it all started

It all started with the extra time I got for a few weeks and the internal newsletter invitation to join the Hackathon. My wife and daughter went to our parents right before the Hackathon start-off date, so I was left entirely alone with our cat. What else can a man do when he’s left alone at home with a cat for 2-3 weeks and working for an IT company? Sure thing, he’d spend most of his time doing work and participating in the Hackathon if there was one nearby. And that’s what I did.

This was the first Hackathon I had participated in, and at the very beginning, it was like, ‘Meh, ok, let’s see what they’ve got.’ So, we teamed up with another Product manager (Alex Goumans) in a sort of Get Rich or Die Trying attitude and embarked on a journey.

Approach

Basically, we went through all of the product development steps in days instead of weeks or months. And I think this was the most exciting thing for me — the speed. We had to accelerate all the phases of the product development processes and squeeze it in a week or so.

• We made ChatGPT-assisted ideation and quickly prioritized the ideas based on the current context and EPAMs business goals. Without going into much detail, we decided to automate everyday tasks that salespeople do using ChatGPT and automation in our existing environment.
• Then we jumped into the discovery and spent a couple of days talking to potential customers/users (from EPAM) to see if there is value for users and the company.
• We did a sort of feasibility prototyping using ChatGPT, prompt engineering, and exploring available solutions to figure out how to connect the dots and make the envisioned solution work.
• Finally, we came up with the mockup design for the app (I love Balsamiq, by the way, especially the ease it lets you create clickable mockups).

By the end of the discovery phase, we still didn’t have someone who could do the job and not just brag about, build a vision, and draw mockups; I mean developers. Our excitement alone wasn’t enough to get the thing working. Thankfully, we got front-end and back-end developers onboard and, together with them, decided on the architecture stuff and briefly prepared user-story mapping for us and the development team to follow through.

And we started building the thing…

The Pitch

To make it clear and strip off the thrill of ‘Is it alive?’. We couldn’t make this thing work by the end of the week. “If only we had more time…” — this is what our eyes were telling silently.

Every team was given around 5 minutes for everything (pitch, demo, Q&A); thus, our story had to be very concise. Not all the teams had a working prototype in the end, but some of them did, and it was remarkable given the fact that they had the same time constraints as everyone else. I think that having a defined problem and potential business impact articulated in a clear and concise story were crucial components along with the working demo.

Nevertheless, our team had a problem to solve and a great pitch with some demo in the end. 5 min of glory, and we went on to wait for the next day to face our destiny.

Results and next steps

Frankly, when the committee started to announce the winners and named our team, I thought we got 1st place, but the winners were reported from the end (ha-ha), so we eventually finished 3rd. And it was awesome! Inspiring moment as it proved that there was a business problem behind it.

Even though we were Inclined to continue exploring the idea further despite the results, this achievement gave us even more motivation to continue working on the project. The next milestone is to build an MVP and get feedback from real users to see how it suits business needs in real life.

What are the key learnings?

Alex Goumans, the product manager whom we worked with along with our developers, has phrased and put some of our key learnings in the final pitch deck.

• Deploy tech faster!
• Be an AI practitioner to develop your mental model of this fast-moving space!

We might also add ‘Engage developers as soon as you can to decide on the architecture or find out key tech constraints.’

Anyway, the whole thing was valuable and fun. I learned a lot of new things in the AI space, even though we barely scratched the surface. We got so much more info and use cases for LLMs, ChatGPT, and AI Agents from the product perspective.

Why you should consider Hackathons (maybe)

Well, it’s up to you to decide whether you want to join if you see a Hackathon. It was a refreshing, engaging, and fun experience for me from a personal perspective, plus I got to know new and interesting people.

If you are a product manager, then it might give you the opportunity to pass all the product development stages in a week or so and find new ideas for your product on the job. It may be your next big thing in the company you work for. Who knows?

If you are a developer, it also can impact your career in several ways, like working with new tech, trying new frameworks, or just talking to a couple of product managers psyched by the idea and the new horizons.

If you are… well, you got the idea, right?

Acknowledgments

Team: Alex Goumans, Ivan Mikhnyuk, Pavel Kiadrynski.

Guys, thanks for this awesome Hackathon!