As a Product Designer and UX Researcher who sits specifically at the intersection of regulated financial systems and LLM interaction design, I was immediately drawn to this project. The challenge of mitigating contextual misuse in AI—particularly in high-stakes domains like finance and healthcare—requires not just design intuition, but a rigorous, structural understanding of how models behave. I bring a unique combination of deep technical literacy in GenAI (OpenAI API, CrewAI, DeepSeek V3) and practical experience designing for the strict constraints of the FinTech sector. Most recently, as an AI Interaction Researcher at the University of Michigan, I moved beyond standard "prompt engineering" to rigorously evaluate model behaviors. I authored and iterated structured prompt frameworks to analyze how different designs shape creative outcomes and potential biases. I designed multi-agent interaction flows (using CrewAI) to simulate complex conversational tasks, documenting specific workflow constraints necessary to ensure user safety and system reproducibility. This direct experience in defining the "rules of engagement" for LLMs aligns perfectly with your need for developing scoring rubrics and designing red-team prompts. My background in regulated industries provides the context necessary to assess risk in "high-harm" scenarios: In FinTech/Asset Management: At Boyd Watterson Asset Management, I designed analytics platforms for government and financial usage. I am intimately familiar with the need for precision, clarity, and policy compliance in UX patterns. I understand that in these fields, a "hallucination" or a dark pattern isn't just a nuisance—it’s a liability. In Quantitative UX: My work has always been data-driven. Whether building portfolio optimizers that reduced analysis time by 60% or running A/B tests on funnel optimization, I approach design with an evaluator’s mindset, backing decisions with metrics and clear rationales. I am eager to apply my experience in establishing rubrics for AI interactions to help your team!…As a Product Designer and UX Researcher who sits specifically at the intersection of regulated financial systems and LLM interaction design, I was immediately drawn to this project. The challenge of mitigating contextual misuse in AI—particularly in high-stakes domains like finance and healthcare—requires not just design intuition, but a rigorous, structural understanding of how models behave. I bring a unique combination of deep technical literacy in GenAI (OpenAI API, CrewAI, DeepSeek V3) and practical experience designing for the strict constraints of the FinTech sector. Most recently, as an AI Interaction Researcher at the University of Michigan, I moved beyond standard "prompt engineering" to rigorously evaluate model behaviors. I authored and iterated structured prompt frameworks to analyze how different designs shape creative outcomes and potential biases. I designed multi-agent interaction flows (using CrewAI) to simulate complex conversational tasks, documenting specific workflow constraints necessary to ensure user safety and system reproducibility. This direct experience in defining the "rules of engagement" for LLMs aligns perfectly with your need for developing scoring rubrics and designing red-team prompts. My background in regulated industries provides the context necessary to assess risk in "high-harm" scenarios: In FinTech/Asset Management: At Boyd Watterson Asset Management, I designed analytics platforms for government and financial usage. I am intimately familiar with the need for precision, clarity, and policy compliance in UX patterns. I understand that in these fields, a "hallucination" or a dark pattern isn't just a nuisance—it’s a liability. In Quantitative UX: My work has always been data-driven. Whether building portfolio optimizers that reduced analysis time by 60% or running A/B tests on funnel optimization, I approach design with an evaluator’s mindset, backing decisions with metrics and clear rationales. I am eager to apply my experience in establishing rubrics for AI interactions to help your team!WWWW…