A look inside the 2nd annual Botpress Hackathon
We’re big fans of hackathons at Botpress. Not only do they produce some pretty awesome results. But for a fast-growing team like ours, hackathons are also a great chance to bring employees together who may not regularly collaborate.
And just like last year’s event, this year’s hackathon (held last month) did not disappoint.
For this year’s hackathon, we focused on solving several known bugs listed on the competition’s Linear dashboard. We split participants into five groups of four (all selected at random). A senior Botpress developer led each team who was in charge of guiding the overall solution approach of their team.
Teams were then instructed to pick any incomplete task from the Linear dashboard to go and solve — no two teams could choose the same task. All tasks were assigned specific points (varying depending on the complexity of the task). At the end of the competition, the team with the highest score would be declared the winner.
This year’s hackathon started at 8:00 am EST and ran a full 24 hours. I had a great team and was very lucky to work alongside a diverse team, including Thomas, Sylvain, and Blake. Thomas was a newcomer to Botpress at the time of the hackathon (he now works on our cloud platform). Sylvain is Botpress’ CEO (and probably created some of the bugs on purpose to win extra points ;) ). I was fortunate to work on Sylvain’s team, given his deep knowledge of the codebase and how the product should behave by default. Lastly, our team included Blake, a talented expert in the NLU system.
My team’s strategy was to begin tackling problems with higher points from the very start to gain more points early on. This strategy paid off and helped us acquire many points before the tasks were all taken — which happened just 12 hours into the competition.
At the end of the hackathon, each team had successfully created an integration branch using all the tasks listed on Linear.
After the hackathon, it took a few weeks to integrate all the tasks into the main branch and test all the features. Next year, we’ll probably consider giving additional points to anyone who adds a unit-test, which would make the code more robust.
I want to thank all of this year’s participants and Botpress for organizing another fun hackathon. It’s such a great way to engage our employees and advance our product. A special thanks to Yann for testing and merging all the tasks and Eff and Jean-Francois for organizing the event!
This article explains by means of an example what a deep neural network is and how they work. Artificial Intelligence (AI) is a broad set of computer science techniques that allow a computer to imitate human intelligence.
Wondering which languages ChatGPT supports? Discover the answer in this useful guide.
With Botpress v12 and Smooch (now known as Sunshine Conversations), it's simple to build a chatbot that you can talk with over WhatsApp. Watch our video!