✮✮✮✮✮

WINNING PROJECT (2022)

✮✮✮✮✮

Inspiration

We were inspired by the commitment, talent & teamwork required to compete in hackathons, let alone win them. DevPost's faithful tracking of attendees & their respective accomplishments reminded us of student-athlete records in the US - often used by regional sports scouts to identify and acquire emerging talent.

What it does

DevRush identifies and connects the world’s most talented hackathon competitors with early-stage startups.

How we built it

  • Data Collection: using Python, we built a spider to scrape competitor data from completed hackathons devpost.com and parsed it into JSON dictionaries.
  • Data Presentation: We built a custom javascript frontend to allow the filtering and sorting of hackathon competitors - as well as view an individual's projects and portfolio links.

Challenges we ran into

  • Data Cleanliness: many DevPost accounts are private, deleted or lack images which can break workflows without exception handling.
  • Runtimes: even with multiple terminals (workers), the performance of Python's requests package is a hindrance to rapid data collection. For our Minimum Viable Product (MVP), we collected 682 of IC Hack's past competitors, but with a longer execution can collect over 100,000 competitor records from DevPost.
  • Cookie Handling: DevPost redirects lazy-loading requests that do not have persistent cookies.
  • Frontend Filtering: React's filtering function is challenging to implement.

Accomplishments that we're proud of

  • We were able to use a powerful insight to develop a practical and useful business solution.
  • We were able to identify and create value within an undervalued market.
  • We developed and deployed a functioning Minimum Viable Product, with a well-designed front and backend, within 24 hours.
  • We resolved all of the major technical challenges that we encountered.

What we learned

  • It is best to start by collecting small batches of data, and layer up complexity when development milestones are achieved.
  • It is best to flag technical challenges early to the team in order to tap into potential solutions.

What's next for DevRush

  • Collect and display the full DevRush platform and integrate into the frontend.
  • Implement time-filtering to filter by recency of a user's activity.
  • Collect data from users' GitHub links to show how active a developer is.
  • Approach startups with the proposition to find high-value, highly flexible programming talent.
  • Connect with headhunters: once developers enter employment, most of their new work becomes "invisible" to the outside world. Our tool empowers talent spotters to approach top talent who are already in employment, to offer potentially better opportunities.
Share this project:

Updates