AI Output Evaluation Specialist (as part of Investment Advisor role at Santander Brazil)
I evaluated AI-generated financial advice for logic, suitability, and regulatory compliance. My work involved identifying flaws and hallucinations in LLM outputs within the finance domain. I wrote adversarial prompts to test AI reasoning about portfolio strategy and risk management. • Applied fiduciary standards to assess AI recommendations. • Rated asset allocation and multi-asset portfolio construction outputs for accuracy. • Ensured AI responses met regulatory and client-specific requirements. • Generated high-quality test and evaluation data for next-generation AI financial systems.