Research focuses on scalable methods for causal effect estimation on industry‑scale observational data—covariate balancing, distributed inference, and communication‑efficient de‑biased estimation—with theoretical guarantees and end‑to‑end deployment on large data platforms, including Snap Inc.