Sohadra Upadhyay

Designed and implemented a scalable platform to understand customer sentiment towards proactive personalized content. The platform is capable of using Machine Learning models as well as heuristic based models to determine whether a specific content is appropriate for the customer or not. The platform is capable of running experiments using different models. The platform produces key metrics and analytics required to measure experiment performance as well as to keep track of customer sentiment over time.

The platform is implemented as cloud based service (AWS) that can be queried through REST APIs. The service code is implemented using JAVA. The service is able to handle 8000+ transactions per second. With the first iteration, the platform reduced customer dissatisfaction by 15.4%

As part of the billing team, my focus was a software module that helps customers (Hospitals) deal with denied insurance claim. It helps customers catch and proactively resolve their denials to help them get payments faster. It also helps report these denials. This lets hospitals identify trends which in turn helps them prevent denials much earlier in the billing process. I worked on a standalone proactive module that lets customers catch problematic pieces on insurance claims before they are sent out the door

I worked on a web portal that aims at making the process of obtaining a loan (Home loan, car loan, credit cards, etc.) much simpler. The aim was to eliminate the need for manually visiting banks and make the experience completely online for end users.