Data / AI Sprints Case Studies
Algorithm and code prototypes
Question Classification in Data Science Sprint with Beagle Learning
ndustry Partner: Turner, MIT EECS, Founder / CEO of Beagle Learning (a question-oriented edu platform for teen and college students)
Story by Dominik: The challenge at Beagle Learning is their outlandish algorithm and unlabelled data. GVC students came in with the “can-do” attitude. They abstracted from NLP, as a rule-based algorithm, to group questions based on hard-coded rules and tagged them with category and keyword. By the end of the event, the team finished their project as a Finalist at GVC and a solution that is ready to be integrated into the Beagle Learning APP. Turner couldn’t be more satisfied with them.
Team members: Araz, Dominik, Bagyasree
Improving Evaluation of TV Viewership through MediaTech in collaboration with Hybrid Company
Industry Partner: Lukas Hnilicka, CEO & Chairman of Hybrid Company
Story by Nathaly: The team worked on a global big data project aiming at enhancing the evaluation of TV viewership.
This included building the MVP, creating 7 reports and visualizing case study scenarios in Tableau reporting and designing views for a presentation.
The team utilized a “raley” strategy with each one taking one task and integrating the individual results into a group work during our cross continental meetings. With one focusing on loading an transforming 3 sets of raw data over 1 million records, and the next student turn data into features. And the last student created reports to visualize data into case studies and prepare for the presentation.
Members: Nathaly, Zuzana, and Roshan