Professional profile

About

My work sits in the validation layer of autonomous systems: turning road anomalies into replayable scenarios, comparable metrics, and regression workflows that can keep running.

How I work

I usually start with anomalies or long-tail cases in driving data, define the problem boundary and reproduction conditions, rebuild the behavior through replay or closed-loop simulation, and validate the change with system-level metrics. The output is not only a conclusion, but reusable scenarios, metrics, and regression entry points.

  1. 01DiscoverFind a problem worth explaining
  2. 02DefineSet boundaries and conditions
  3. 03ReproduceReplay or closed-loop simulation
  4. 04ValidateMetrics, comparison, regression

Published here

The site separates information by maturity: long-form methods, curated case studies, and lightweight updates.