Interesting idea. Comparing cities is usually scattered across a bunch of sites. What data sources are you using for things like cost of living and quality of life?
Agree with this direction. Most tools assume structured inputs, but real workflows are messy and visual. Feels like screen understanding is still very underexplored compared to text.
This is actually interesting. Feels like we’re moving from “generate UI” to “validate UI,” which is a completely different problem. Curious how you handle edge cases where something looks correct but breaks in interaction?