Designed a multi-platform predictive AI/ML analytics solution in close partnership with a cross-functional scrum team of over 15 engineers, shaping product direction and aligning design strategy with technical implementation.
Led user interviews and collaborative discovery workshops to surface workflow needs and gain consensus among PMs and executive stakeholders, ensuring the product vision was grounded in real operational insight.
Delivered fully custom, high-fidelity designs for logistics management flows and graph visualization experiences, guiding iterative design reviews to refine usability, clarity, and interaction quality across the platform.
The client was facing a dilemma of lack of data connectivity in a complex operational environment that involved multiple stakeholders and data owners that needed to be able to communicate and effectively transmit necessary data elements to one another in a secure environment.
building trust and focusing on the problems to solve
I facilitated cross-organizational workshops to define the problem space, map the data landscape, and identify requirements. The goal was also to build trust with these stakeholders who both owned the relevant data sources and housed the potential future users of the product. It was crucial to guide the multiple stakeholders with varying interests towards the same priorities. In this process, we also gained an understanding of how information flowed and where silos were hindering communication efficiency.
DOCUMENTing A SINGLE SOURCE OF TRUTH
For most projects, I believe in documenting a single source of truth for the existing state for everyone on the project to reference as we move forward with design, development, and continual client engagement. It is critically important for keeping everyone aligned on the ”WHAT” and the “WHY” of a project.
KEY INSIGHTS: Problems to solve
Data sits in disparate siloes and all coordination between stakeholders happens over email.
When customers request information, there are too many steps involved in reaching out to various parties for the relevant data.
Customers do not have proper estimates for delivery.
Stakeholders cannot easily track the responsible party when mistakes are made.
DESIGN DECISIONS
Application of color to indicate data source and stakeholder responsibility.
Visual emphasis on duration of time rather than date.
Each process can expand to allow the user to explore and investigate each step.
Only one process can expand at a time to prevent information overload.
UX OUTCOME
User can easily visualize package hand offs, delineate responsibility between stakeholders, and explore each node progressively while still understanding the context of the overall process.
DESIGN DECISIONS
Consistent color designations carried over from timeline view.
Labeled nodes and relationships to provide more context.
Information panels are consistent across the visualizations for each event node.
Predictive information for in progress events is clearly displayed in panel.
UX OUTCOME
User can easily visualize relationships between events, delineate responsibility between stakeholders, and explore each node progressively while still understanding the relational context of the data.