Inside Devpost’s Internal Hackathon: Exploring AI With OpenAI

Innovation is core to Devpost’s culture. It’s not just a buzzword, especially in today’s AI-focused tech landscape. It’s never been more important to experiment and look for creative solutions. So, we did what we know best—we ran a hackathon! 

Our engineering department hosted an internal AI hackathon using the Devpost for Teams platform. Here’s a look at how we ran it, what we enjoyed the most, how it was valuable to the organization, and what’s next on our roadmap.

Goals for the Event

The hackathon focused on exploring AI’s potential, particularly the OpenAI API. We aimed to integrate AI-driven features into our platform to benefit our users. 

The hackathon was part of the engineering team’s onsite, so we also saw it as an opportunity to collaborate in a new way. During the 2-day event, we hoped to make the most of this short time away from our day-to-day work and come up with new creative ideas.

What Did We Do?

During the hackathon, we developed features that seamlessly integrated with our codebase. These features included generating project ideas and creating images for projects based on user-provided prompts. We also used AI to enhance project descriptions and criteria for hackathon submissions.

Mastering the art of crafting effective prompts for the OpenAI API was a challenge. We spent time reading documentation and watching training videos to become proficient in this area. This learning process was a valuable foundation for our work during the hackathon.

What Did We Enjoy the Most?

The most enjoyable aspect of the hackathon was the opportunity to dive into AI. The entire team relished learning how to leverage AI to produce valuable results. 

Collaborating on this shared learning journey was equally rewarding. We tackled challenges together, brainstormed solutions, and shared knowledge—fostering a sense of camaraderie.

How Was It Valuable to the Organization?

The hackathon brought concrete benefits to our organization. Our engineers acquired new skills and knowledge, enhancing our team’s capabilities with AI. 

Additionally, the AI features we developed sparked ideas for our product team. Since we have working prototypes coming out of our hackathon, these ideas can already be discussed in customer feedback sessions, and we're excited about the possibility of developing them into new features for users. 

The experience also provided insights into challenges associated with AI integration. The biggest of these was the classic problem of GIGO (Garbage in, Garbage out), where we found that if we didn’t prompt the AI model well enough, the output it gave us was essentially useless. Having the focused time in the hackathon to experiment with providing better prompts allowed us to work through this challenge and others.

What’s Next?

Our AI journey continues beyond the hackathon. We’re currently using prototypes developed during the event to conduct user testing. Gathering user feedback is essential to refine these features and identify areas for improvement.

The insights from user testing will guide our product roadmap. We’re excited about the possibility of translating this feedback into practical product features that will enhance the Devpost experience for our users.

Making the Most of Internal Hackathons

This internal hackathon was a practical exploration of AI’s potential within our platform. It allowed us to learn, collaborate, and contribute valuable insights and features to our organization. As we move forward, we’re committed to harnessing the power of AI to continually improve Devpost, ensuring that we remain at the forefront of innovation in the tech world. Stay tuned for updates from Devpost!

Learn more about the ways your team can experiment with AI with a free demo of Devpost for Teams.