Product Clinic #5: Lindus Health’s first Hackathon

In July 2022 Lindus Health commenced its first hackathon with everyone in the product team, from data science to our CTO and head of product getting involved. 

‍

First things first, what is a hackathon?

Hackathons can be stereotypically misconstrued as a group of engineers in an unreasonably dark room, quite literally “hacking” against the clock day and night or until they succumb to the numerous tequila shots provided to them (disclaimer: this is a reference to a scene in The Social Network, there were no tequila shots at the Lindus Health hackathon, at least not this year). 

What a hackathon really is, is an opportunity to let ideas and collaboration run free over the course of a few days. The core software engineering principles of building maintainable, well written and well tested code are temporarily replaced with a focus on creating something new and innovative. The work built is ideally related to the core product base and may even have potential to be a future feature. Essentially, a hackathon allows team members to build whatever they want however they want with the primary goals being creativity, trying something new and having fun. 

So what did our hackathon look like?

The general structure of our hackathon was an opening meeting followed by idea generation (with some fantastic ideas contributed across the entire Lindus Health team). Which led into two days of cross collaboration and “hacking”, with all other work put on pause, at least for the most part. At the end, the work was demoed across the company. Two days, around 20 Crosstown doughnuts and a decent amount of pizza later, the Lindus Health hackathon demos were ready. Despite the short stint of available hours and a curiously timed client facing demo, the output from the entire product team was impressive. ‍

So what did we build?‍

Within the plethora of demos we had a gamified participant leaderboard - whichever participant can complete their trial questionnaire fastest gets the highest score! (which is needless to say, not something we would encourage in a real trial); a live participant status board to observe our product’s real time usage; an SMS integration (coined SatsuMaS) with our web based questionnaires to make our product that bit easier for participants to use and some “very hacky” CSS to produce an affirmation providing lemon (think a very basic implementation of the Microsoft Office clippy but in the form of a citrus fruit) that spontaneously rolls across the screen providing words of encouragement.‍

Naturally data science got involved. ‍

We also tried to train a machine learning model to generate trial titles and see what it would come up with. Based on a list of titles downloaded from clinicaltrials.gov and some open source python libraries we created a model that generated the following ideas for future clinical trials:

• A Study of Positive Endoscopic Application in Patients With Moderate Menolitus
• Effects of Anti-Horuvant Research on Patients With Metabolomia
• Efficacy of Catkle-responses in Perioperatuver

Needless to say, we probably won’t be picking up any of these studies soon.

Last but by no means least.

We created a confetti effect across our dashboard, the future use case of which has not yet been pinned down, but which serendipitously brought the 2022 hackathon to a close. It is safe to say it was a huge success and we are all definitely looking forward to next year.

‍

Madeleine De Forest-Brown, Senior Software Engineer at Lindus Health