README

The Product

Everyone knows the name of the game. Brave the endless coffee chats, the application, multi-round interviews (run by people no more than two years older than you) to get into the right club, work your way up to Executive Board by any means, network at every turn, and soon you'll be looking at an offer letter from the company of your choice. The not-so-easy part is choosing which club to spend 10 hours a week for the next 4 years.

At its heart, Elitist is a centralized database that tracks where alumni from Cornell's engineering and finance student organizations end up working at. Normally, all this information is scattered across Reddit posts, word of mouth intel during Clubfest, and self-congratulatory club info sessions.

The Data & Methodology

While I considered web scraping LinkedIn to get a more thorough insight of the placement data, it's anti-scraping detector made this idea all but impossible. Instead, I used the Alumni search future on Cornell's public LinkedIn page as this allowed unlimited searches, which proved useful as 50 clubs with 50 companies each would mean at least 2500 searches. After some refinement, I found that "club name" in parallel with the "where they currently work at" filter produced the least false positives and negatives.

The second setback I ran into was the data itself. Except for several outliers, the numbers were mostly 0s, 1s, and 2s (which were how many people work at different companies from a particular club), which meant that the data wasn't that helpful. I solved this issue by batching the largest companies into an industry, and calculating placement based on the industry itself, which helped smooth out the noise and provided a broader view of the dataset.

Finally, other data such as time commitment per week and recruiting schedules were obtained directly from club leadership.