Hehe, posting this because I wanted to show it off.
The concept is to collect holiday data from Korea, Japan, and China along with historical visitor statistics to predict and score how crowded a place will be.
It combines rule-based logic with LLM analysis, and for the most recent 30 days it runs daily LLM analysis so that qualitative predictions are included as well. It also recommends an optimal route based on whichever pass you purchase.
I went ahead and implemented all of that.
It's been fun. If you have any feedback at all, I'm always open to hearing it~
https://hansanday.com
