Independent researcher.
Systems builder.
Evidence first.

I am based in Qingdao, China. My trajectory spans finance and econometrics, database technologies, multilingual NLP, AI safety assurance, and end-to-end digital systems. That range lets me move between empirical analysis, technical implementation, risk reasoning, and operational decision-making.

Completed work includes a multilingual system for toxic and xenophobic language detection, an AI Safety Testing Framework covering standards mapping, threat modelling, adversarial testing, multidimensional metrics, and mitigation validation, and quantitative modelling of interest-rate behavior using ten years of Chinese bond-market data.

Alongside research, I lead product and operations work for a B2B digital platform spanning a WeChat Mini Program, back-office system, website, backend services, deployment, and workflows across five operating domains. Current agent-safety research is listed publicly by topic and status only; its methods, evidence, and manuscripts are not disclosed.