Invesco is an 8,000 person, global investment management firm primarily serving institutional investors and manages upwards of $1.5 trillion Assets Under Management (AUM).
• Through strategic product direction and user-centered design, I helped lead the evolution of IVYGPT from a basic chatbot into Invesco’s core GenAI platform.
• I shaped the feature roadmap around user insights and business value, ensuring clear differentiation from external competitors while aligning closely with senior leadership and firm wide AI priorities.
• My efforts contributed to a 24-point increase in NPS within 3-months and facilitated its growth into a competitive tool despite limited team resources.
• I led the discovery, scope, and design of a high-impact GenAI tool that automated custom model commentary, a regulatory-heavy and resource-intensive task.
• I leveraged use case models pioneered by McKinsey and Amazon and other design thinking models to determine scope and focus on maximum business impact.
• The tool enabled Invesco to scale from $2.2B to $9B in custom asset models without new hires, reduced quarterly production timelines by over 7 days, and required zero UI changes in its first 18 months of use.
•The project was cited as our team’s most successful internal launch to date due to guardrails, scope, and persona constraints I defined early in the project that kept scale bound by end goals and within regulatory boundaries.
• As lead designer and AI design thinking lead at Invesco, I identified a pattern within low-impact, document-related GenAI use cases and a proposed scalable alternative.
• I designed a modular document interrogation tool that allowed users to upload documents, build and manage custom prompt groups, and share workflows without developer support.
• Despite infrastructure challenges across four development squads, I stepped in to coordinate execution, maintain momentum, and realign priorities.
• The result was a scalable internal tool that addressed 80% of related use cases and increased operational efficiency without needing the core team of data scientists to make consistent changes.
•It empowered users to test and validate their own workflows thereby freeing the core AI team to focus on higher value use cases requiring more bespoke solutions without slowing GenAI adoption at the firm.
