top of page
3 happy developers collaborating in front of a shared screen.

Case Study: Enterprise ML Web App

Revolutionized project transfers, increasing by 42% and slashed loan defaults by 30% for Fortune 500 clients by crafting a reusable framework.

Industries: Data Analytics, Enterprise and Financial Services

Platform: Web

Competencies: Research partnerships, interviewing, competitive analysis, documentation, training, data analysis, prototyping, design systems, Agile development

Team: 2 Senior User Researchers, Project Lead, 4 Data Science Researchers, 2 Business Analysts and 3 Developers

As the Principal Interaction Designer, I cultivated design innovation for emerging technologies at the Xerox Innovation Group (XIG), a SaaS B2B research organization with 4 international centers, including Palo Alto Research Center (PARC).

In 2012 I faced the challenge of an increasing number of data analytics projects that lacked design support, resulting in low success rates transferring proven research concepts to business units for productization.

Image 1. Cover page for "User Experience for Analytics: Design Guidelines".

Cover page for "User Experience for Analytics: Design Guidelines." Created by Zahra Langford, Jennifer Englert and Pat Wall in January 2013.

Task

My mission was to rapidly gain expertise in data analytics and machine learning (ML), then create a reusable design framework to extend design support across multiple teams, ultimately improving project success rates.

Actions

Research and Document Framework

  • Collaborated with the user research team to create the "User Experience for Analytics: Design Guidelines" (UXADG) framework, spanning 44 pages.

  • Planned and participated in a comprehensive research study with two senior user researchers, we interviewed and observied12 internal algorithm specialists, 14 external knowledge analysts and 3 internal analytics tool development teams.

  • Completed a competitive analysis of seven popular self-service data analytics tools and Microsoft Excel.

  • Integrated research findings into a simplified design process, ensuring ease of use for non-designers.

  • Provided training to 6 teams on the framework and served as a consultant to support their adoption.

Use Framework on New Project (Xerox Adaptive Risk Management - ARM)

  • Simplified the utilization of machine learning (ML) and predictive analytics for internal Business Analysts, supporting their enterprise clients in financial services through the ARM platform.

  • Interviewed two business analysts and collaborated with the project team leader to define project requirements.

  • Led team brainstorming and whiteboard sketching sessions, iteratively developing the information architecture and wireframes.

  • Directed developers to utilize components from Xerox's design system, collaborating with design system owners to add new patterns for complex data analytics workflows.

  • Maintained close collaboration with offshore and onshore developers throughout the development and delivery phases.

Results

-30% Loan Default Rate +42% Project Transfer Rate 100% Data Analytics Project Use

  • 100% of data analytics project teams used the framework in their processes in 2013.

  • The UXADG framework and training increased project transfers by 42% (from 0/3 to 3/7).

  • Framework used in design of ARM platform used to slash loan defaults by 30% for multiple Fortune 500 bank pilots.

  • ARM platform transferred to business unit and offered as a Xerox Financial Service capability.

Solutions screen from the XAR platform showing completed results of a borrower risk analytic and associated actions.

Image 2. Solutions screen from ARM (formerly Multi-Application Predictive Analytics) showing completed results of a borrower risk analytic and associated actions. 

bottom of page