Reimagining a PMR tool (Promotional and Medical Review) which is a ML based B2B application that simplifies and speeds up reviews & approvals of promotional medical materials. The client is one of the largest Biotech firms in the world serving millions of customers across the globe and this particular offering was targeted toward the US region. 

Objective - Redesign and Revamp the application to increase task completion rate & efficiency through ML automation

Pain points - (5 - 10)% of total FDA submissions are late, 30-40% of agency submitted files do not follow submission guidelines, 50-80% of errors are preventable. Apart from that there was a lot of negative feedback about usability and understanding of the features. And all this data was collectively provided by the client themselves.

My Role (Claims Module) - User Research (interviews, desktop research), Documentation, Heuristic Analysis, Wire-framing, Prototyping, User Testing, Visual Designs

Team Size - 12 

End Users - MLR (Medical Legal Review) teams or experts. They have background in scientific or drug related/medical domain

Step 1 - Understand Business requirements, domain related business logics, pain points in the existing flows

Step 2 - Many work-flows were redefined based on the real-time constraints in the market rules and real data. This helped define scenarios and its related use-cases

Step 3 - On an iterative basis, the understandings on each feature were validated over a collaborative tool and the concept mocks were finalised based on early usability related user/stakeholder feedback. The good to have features, technology dependent patterns and interactions were documented for future scope

Step 4 - Detailed visual designs scenarios with edge cases were crafted for most popular desktop resolutions and shipped for development and last minute design updates incorporated

Step 5 - Developed screens are reviewed/audited for bug fixes and data sanity

Results - This product was successfully able to reduce the overall turnaround time by over 30% and saved around 2500+ man hours within a few months of implementation which is like a massive improvement.