You don’t have time to organize a hackathon? Think again.

Technology inevitably speeds up, and your meeting calendar is not getting any less occupied. Yet, you know you have to be innovative to keep up – for your own good and the company’s too.

This is the reason why Infobip started regularly organizing hackathons. Since then we now have a steady stream of innovations trickling downstream.

One of the hackathons we organized was devoted to Generative AI. I’ll tell you how it went.

2023 was the year the Generative AI market exploded, leaving everyone bewildered. Everyone wanted a piece of the pie: companies scrambled, and GPUs were sold by the shipload.

But, despite the hype, there was no clear picture of how to integrate this non-deterministic technology into a primarily deterministic world of software engineering. And we had to find a way to create concrete use cases for the technology. 

But, we faced one big problem: Developers are being left at the mercy of their schedules.

This is why we introduced the idea of a hackathon – an event that can give us a time, a place, and someone to support the builders amongst us to implement this new technology. Or, in other words, to streamline the process of integrating this new technology into their “tech stacks”.

What is the first thing if you want to organize a hackathon? You have to find time. While this may seem easy, believe me, it’s not. 

We already said – that developers are slaves to their tight schedules. 

This can be said for any department of a typical ICT company, but even more so, this can be applied to the engineering department. 

The truth is that engineers like to have everything neatly organized, from task lists, and meeting agendas to the way of work. And that is completely fine but it can constrains us from being innovative. We should not fall into the trap of following the anti-pattern of blindly following the delivery through seamless execution. 

Building a house without someone to use it doesn’t make much sense. The same can be said for hackathons, and that’s what makes them dangerous events to host if you haven’t planned everything out. 

Remember, we’re dealing with IT folks here and nobody likes to venture into something that doesn’t resonate with the opportunity or a business purpose. 

This is why the theme of the hackathon should solve some specific business problem, or at least enable the folks to explore some areas that are of strategic importance to the company. 

This is why we first chose GenAI as a topic, but even why we had to organize a second iteration to tackle all the great ideas our engineers came up with.

Having to deal with GenAI also puts you in the most difficult position because the technology is so bleeding edge that it’s hard to get a feeling for that very same edge. Generative AI usually feels somewhat like black-box magic to anyone outside of the machine-learning space so the opportunity and purpose must be communicated. As must be the preparations and support for this event to succeed. 

Also, the marketing perspective is very important. 

The idea of having an event that nobody knows about is worse than having a most popular one, and you want to be heard among developers. This audience is not just your innovators but also builders who can make things work and bring big tangible value and do it quickly.

To resume, you want to be heard, and for engineers and developers, there is no such thing as over-communication. 

Sure, someone can put you on their spam list or put a personal firewall to your ideas, but those people are usually not your audience anyway. 

Supporting people on-site is also mandatory for the hackathon to play out well, especially if you are venturing into some area of IT that is not yet that popular or well-explored. 

In our case, it also meant that we had to be on-site for all the support in providing infrastructure for the GenAI projects and both providing some best practices for Retrieval Augmented Generation, Prompting, and other cases so it was also a learning opportunity for all involved. 

Having people deal with basic setup procedures (think permissions, deployments, authentication, etc.) just puts everybody off and you fail. 

When thinking about supporting people around GenAI we resorted to having a beefy resources list with: hello world examples in different languages, model deployments ready (both smaller and larger models…you never know the use case), and basic terms and good practices. Engineers like to prepare, and we need resources for this.

As a company, you may wonder if it’s truly worthwhile. Hackathons are not easy to organize; they’re time-consuming (and time is money, after all), and the results are unpredictable. However, one could argue that innovation often thrives under pressure.

Nonetheless, innovation, engagement, and knowledge sharing rank high on most companies’ wish lists. Hackathons serve as platforms where ideas evolve into tangible solutions that add value to customers.

Participating in a hackathon is a hidden gem for both participants and organizers, as it facilitates experiential learning, particularly in software engineering. In our experience, the momentum generated during a hackathon often extends beyond its conclusion. 

We not only successfully introduced the technology but also ignited a drive among our engineers to further develop their skills in Generative AI – a crucial step for any company considering its adoption.

Then at the end of the whole thing, you will find yourself with a bunch of cool projects that are in desperate need of feedback. 

In this part, it’s a good practice to include folks from Product Development, since you want the best ones to end up in a product, either internally or externally. We’ve observed a roughly equal distribution (50-50) of success stories between internal and external projects. Ultimately, through a single hackathon, we’ve witnessed Generative AI both enhancing our products and assisting us with day-to-day operational tasks 

In my opinion, there’s no better way to introduce engineers to new technology in an organic, seamless, and successful manner. Therefore I strongly encourage you to try it out.

To conclude, I’ll say it again: in today’s market of exhausted ideas and stagnation, the most innovative reap the rewards of plenty.