Timeline
48 Hours (25 April - 27 April)
Tools
Python, R, Pandas, Matplotlib, Shiny, Figma, Tableau
Team
Daniel Wang, Hannie Xie, John Tan, Ming Yang Li, Oscar Su
🏆 Winner: Judges' Choice Prize
The Challenge
Real estate companies hold vast amounts of data, but often struggle to identify which regions hold the highest potential for future growth across different sectors.
⚖️~75%
of leasing decisions made without predictive analytics
💸$20-60 million
in annual revenue lost
🕰️6-9 months
wasted scouting in low-growth areas
💡 Fueled by adrenaline, and lots of snacks and caffeine, our team developed a solution to tackle this challenge—all within a whirlwind 48 hours.
The 48-Hour Process
Fri 2/25 9 P.M.
Sat 2/26 12 P.M.
Sat 2/26 11 P.M.
Sun 2/27 10 A.M.
Discover
- ・ Brainstorm
- ・ Research
- ・ Data Cleaning
- ・ Exploratory Data Analysis
Define
- ・ Research Question
- ・ 5W1H
- ・ Workflow
Develop
- ・ Data Visualization
- ・ Modelling
- ・ Interpret Results
Deliver
- ・ Practice
- ・ Presentation
- ・ Pitching & storytelling
❓Questions we asked ourselves throughout
(Our Mid-Way Breakdown: Hit a wall 24 hours before the deadline, scrapped our plan, took a walk, downed more caffeine, had a deep 1-hour discussion, and came back with one clear goal.)
Putting everything together...
Before: Densely plotted map
After: Map with Sector Markers
Before: Hard to interpret
After: Improved readability and Insight
🌟The Product🌟
Final Thoughts
Sleep? Not at DataFest. It was a blur of caffeine ☕️, chaotic brainstorms, and surprisingly productive late-night breakthroughs—and I loved every second. As my first hackathon, it taught me how to turn messy data into something actually useful (and maybe even a little pretty).
I walked away not just with a working solution, but with a much deeper appreciation for the creative, scrappy process of data storytelling. Big shoutout to my teammates for being insanely talented, and the best people to be stuck in a 48-hour data sprint with. Would absolutely do it again… after a long nap.
