At WorldQuant BRAIN, I researched, validated, and deployed 140+ active production alphas across global equity universes, with in-sample Sharpe ratios up to 2.96. I directed AI-assisted Python and FastExpr workflows and designed a closed-loop research system with browser/API automation and platform verification.
I also architected a local knowledge base and hybrid retrieval workflow with 273+ research nodes, provenance tracking, and validation gates. My work spans multi-agent orchestration, spec-driven development, and AI-assisted product prototyping.
At Oracle via Scale AI, I evaluated and ranked AI-generated outputs and designed adversarial prompts and rubrics to surface model failures. Earlier, I created and audited computer vision annotation datasets at DiscoveryLoft Inc. Vietnam, and conducted environmental modeling and remote-sensing analysis in research roles.

