LUCID · local causal engine

1 · Connect

Database URI (read-only recommended)

2 · Discover

Root table (auto-joins outward along foreign keys) Target column (outcome of interest)

Causal map

Connect and run discovery to see the causal structure of your data.

3 · Ask

Supported forms: if we increase X by N, what happens to Y · what happens to Y if X drops by N · effect of X on Y · percentages like by 5% work too. No amount = one standard deviation.