Recap from LinkedIn’s Intern HackDay
On July 15-16, LinkedIn held its sixth Annual Intern HackDay at our headquarters in Mountain View, California. More than a thousand college interns from tech companies around the Bay Area applied to attend the 24-hour event, which included a video game lab, a nap lounge, midnight food trucks, an interactive graffiti wall, and tons of pizza! LinkedIn’s VP of Engineering, Feed & Content Products, Erran Berger, kicked off the event Friday evening, and the interns did not waste any time before beginning to hack away!
Winning teams
There were more than 40 teams who participated in the hackathon, 11 of whom made it to the final round of judging. The final judging panel was made up of LinkedIn’s Erica Lockheimer, Figma’s Dylan Field, and Piazza’s Pooja Sankar. The winning team—Support Interns for A Cause, with interns Shawn Ren, Bradley Chee, Mark Ye, Bobby Zhou, and Dao Lu—took home the $10,000 grand prize for their project, CrowdCompute.
Support Interns for a Cause pose with their first place prize.
Here’s how the winning team describes their project:
Our project, CrowdCompute, is a distributed computing platform that crowdsources MapReduce jobs to run in a large number of website visitors’ browsers. Jobs are run in a distributed fashion across multiple clients, taking advantage of otherwise unused CPU power. The original inspiration for this project came from trying to innovate a new type of web economy to replace advertisements and introducing a way for developers to work asynchronously.
The process begins when a developer or company visits the CrowdCompute web client. There, they can submit a request in the form of a map and reduce algorithm (currently only in JavaScript) to execute and the data to process. Once the request is submitted to the system, the data will be split off and each data partition, containing relevant context from the original job, will be scheduled for delivery as a job. When a web browser visits a site where the CrowdCompute plugin is supported, it will pick up the next available job and begin executing the provided algorithm for the provided data and deliver the result back to the server. Once all of the jobs associated with a request return results, the results are then collated, reorganized, and delivered back to the original submitter.
While we were able to present a working demo, there are a lot of areas for improvement. For example, many people raised concerns about security, which we plan to revisit and eventually alleviate through extensive research on industry standards. Additionally, we plan to make the platform more flexible so that it can handle a larger variety of jobs in a larger variety of languages.
Second place in the competition went to the Gold Team, made up of Logan Carmody, George Farcasiu, Thomas Lomont, and Tina Luo, for their project PolitiHack, an SMS chat bot providing updates about important political decisions. Third place went to Team Swish’s Sid Malladi and Kishan Patel for their project Hype, a customer loyalty solution for small and medium-sized businesses that allows consumers to unlock rewards by tracking their engagement via a mobile app.
Acknowledgements
A huge shoutout goes to the Intern HackDay planning team, all of the volunteers, and Emily Krebs for leading the team in executing such a great event.