What we build
We build a nationwide synthetic population aligned to official statistical distributions.
| NAME | SEX · AGE | LOCATION | OCCUPATION | VALUES (10-D) |
|---|---|---|---|---|
| Ahn Seong-eun | F · 37 | Seongbuk-gu, Seoul | Office worker | |
| Oh Eun-hyeon | F · 26 | Icheon-si, Gyeonggi | Student | |
| Jo Yeong-jin | M · 22 | Ansan-si, Gyeonggi | Student | |
| Jang Yun-hun | M · 44 | Guro-gu, Seoul | Manual work | |
⋮ 999,996 more rows · 1,000,000 total | ||||
We simulate surveys (Million Answers) and roundtables (Persona Roundtable).
We put the same questions to real respondents and compare. (My Answer)
Where it is used
Urban planning and public policy, social research and market research.
They share one thing: decisions that need to ask people, where asking is too expensive or comes too late.
How we work
Synthetic responses reproduce real response patterns to a degree. They do not replace them.
We compare distributions against public survey data and calibrate against parallel real surveys.
Outputs carry the sampling method and its sources.
Accuracy drops for new phenomena and contexts absent from training data.
Contact
For press, partnerships, investment, and research collaboration: contact@deepurban.net