Amit Khajuria

Partnered with Product development team to create AI based product by leveraging various machine learning algorithms and analyzing cross sectional/operational data. This helps to predict customer satisfaction score and understand relative impact of key drivers/influencers, achieved increase in Net Promoter Score by 12 points. Developed and deployed patient risk score predictive model for clinics/hospitals using random forest algorithm and analyzing cohorts of patients with similar characteristics, led to decrease in admission rate by 18%. Built patient no-show predictive model using gradient boosting machine technique and deployed in production which helped client to optimize resource utilization and resulted in $0.7 Million in incremental annual revenue.

Implemented demand forecasting technique to predict sale of SKU's over next six months, decreased dead stock by 25% and increase profit by 5%. Built database by integrating different legacy data systems using Star Schema approach and leveraged it to built multiple tableau reports, helping Client to understand periodic trends and optimize claims operation, led to bring down claim cost by 15% and saved $0.6 Million annually. Designed and created tableau reports to provide business with ability to track and review performance of fraud management system, decreasing cost per claim by 3%. Developed and deployed set of predictive models for different nature of motor claims to prioritize claims based on fraud propensity using various claim and customer characteristics. Integrated with tf-idf model scored on claim description to achieve higher accuracy. This led to realization of $1.5 Million benefit annually. Created Property and Specialty Claims Reporting Suite which helped Portfolio Managers to take key financial and operational decisions and shorten claim life by 18%.

Served as a member of Placement Advisory Committee for the year 2015 as a representative of Metallurgical and Materials Engineering at NIT Srinagar.