In 2013, Cybera launched the Rapid Access Cloud (RAC) in order to fill a gap in Alberta research and education: access to free but powerful compute infrastructure to run tests and support learning. Over the last seven years, as usage of our cloud took off, we have been amazed by the breadth of projects we have supported: from public-serving apps, to 3D animations, to a range of AI and machine learning research. But we have always had a soft spot for education, and supporting classrooms to gain experience in the use (and potential) of cloud.
Almost from the beginning, Eleni Stroulia has been one of our biggest advocates of using RAC in the classroom. Stroulia is a Computing Science Professor at the University of Alberta, and for several years, has utilized the RAC for her CMPUT 401 – Software Process and Product Management class.
To add an interesting challenge to this course, Stroulia incorporates a hackathon for her students. This semester’s hackathon took place this past weekend, with teams of 4-6 working for 48 hours on a data problem.
“Their project doesn’t have to be related to their classwork, we just want them to have fun and build team spirit,” says Stroulia. “The most important thing is that they’re getting hands-on experience with RAC. There’s a little bit of configuration, and trial and error, but it’s a great way for them to learn how to deploy things on a basic cloud infrastructure.”
The Top Classroom Projects
A team of four students made up of Cole Merkosky, Joe Bosch, Brian Grenier, and Brett Commandeur decided to create a song recommendation tool based using the Spotify API. As they described: “New ways of song discovery are always in high demand.” They created a list of parameters that included danceability, “instrumentalness”, happiness, “accousticness”, and loudness to create Dalmation, a web app that helps users find new songs based on their unique tastes.
The judges gave particular praise to the Dalmation team for “figuring out how to make use of netlify to enable the deployment of the playlists on the user’s account!”
Another team of four, Sean Coutinho, Isaias Briones, Ohiwere Ahimie and Xiang Fan, set out to find a solution to a very 21st century annoyance: “There’s more content than ever, yet it seems like there’s nothing to watch!”
They created a system that draws on the database of Netflix programs and presents them on a standalone website, searchable according to criteria such as “How many minutes do I have to watch something?”. In their description, the team says they were proud not only of creating their platform in just two days, but also figuring out how to leverage Rakuten API (one of the world’s largest API marketplaces).
Accidents Down South
The third-place team was made up of Khrystina Vrublevska, Abenezer Belachew, Dan Zhang, Jason Lee, Vivek Acharya, and Gengyuan Huang. Their project, Accidents Down South, sought to raise awareness of traffic accidents, with the ultimate goal of showing people when and where they should be more cautious driving.
Using US data on countrywide traffic accidents, they mapped the months and hours that faced the highest number of accidents, as well as geographic areas that reported the most (and the most severe) accidents. The group was proud to report that they were able to work with a very large dataset (2.25 million records, 800 MB in size) to run their analyses.
Building cloud confidence
Stroulia sees hackathons like this as not only a fun exercise that gets students to develop their cloud skills, but also a great way to get them working as a team, and thinking of real-world solutions.
“Last year, one of my students told me that he was thinking of dropping out of the course, but the hackathon changed his mind,” she says. “He liked working with his team, and developing a tool that would have a wider use outside of the classroom. In their case, they built an application that could be used by the University’s student union to list the available activities during the first week of school, and award students who did more activities. It was deployed last September and used by 3,000 people, which was great.”