AI & Data Transformation
As the Lead Product Designer, I spearheaded the design strategy and execution for a multi-language Gen AI-driven Quality Deviations Management system. This initiative involved translating complex business requirements into a user-centered prototype design, significantly enhancing operational efficiency and compliance across six countries. The system is tailored for over 3,000 users and is projected for global release in June 2025.Key achievements include an estimated 30% reduction in average investigation time and a remarkable saving of 87,500 Full-Time Equivalent (FTE) hours following the Minimum Viable Product (MVP) deployment.
Additionally, we anticipate a 25% increase in the average IQS Score within the first year of implementation.My role encompassed conducting comprehensive UX research for the AI Investigation Digital Assistant and the Standard Operating Procedure (SOP) Assistant, which played a crucial role in optimizing product development and crafting seamless AI-driven user experiences.
I facilitated workshops and User Acceptance Testing (UAT) sessions to gather feedback, driving continuous UX improvements.Furthermore, I mentored two designers in the implementation of the design system within the GenAI Value Creation program, ensuring a cohesive and scalable design approach. This leadership not only fostered a collaborative environment but also enhanced the overall quality and consistency of our design outputs.
The tool will assist with suggested technical writing designed to help meet IQS standards, based on the raw data entered by SMEs, and supporting Global Deviations Management for 50+ sites, 3000+ end users by June 2025.
The AI generated output is a first draft of text which must then be reviewed and/or edited by the original SME who entered the data before they take responsibility for entering it into the system of record. The user owns the output content and can reject or modify the content. The AI Investigation Digital Assistant is a standalone business tool and will not be integrated into any Takeda Northstar systems. As the investigation and report writing process is currently manual "knowledge work" it does not impact any existing Deviation methods or the official GxP repository (TrackWise).
The System includes:
In the system, Retrospective Review is the only feature is partially impacted by GxP.
