Shubham Kumar Email & Phone Number

Bengaluru, Karnataka, India

San Francisco Bay Area

• Built two real-time Tableau dashboards to mitigate medication shortages by evaluating medication inventory on hand, usage/transactions and corresponding costs trends to enable forecasting capabilities based off a dynamic historical range• Spearheaded a team of 5 to ensure smooth delivery (dashboard design, user-testing and final deployment) and led weekly client meetings to highlight progress, risks and mitigation strategies• Drafted a project management playbook for UC Davis Health team, listing requirements and necessary tools for future practicum projects

Abu Dhabi, United Arab Emirates

· Built a look-alike model using logistic regression to identify prepaid subscribers likely to switch to postpaid for Etisalat· Helped in identifying subscribers likely to switch to Etisalat app for bills and payment in the coming month by building a machine learning model using the DPI behavior of subscribers· Analyzed the impact of VoIP apps on subscriber usage and revenue during COVID-19 period by identifying potential indirect VoIP app users and creating derived controls

Abu Dhabi, United Arab Emirates

· Part of the Data Modelling Team within the Customer Analytics and Insights Department of Etisalat· Evaluated the performance of various machine learning based marketing campaigns using derived control· Built a random forest model to identify new VoIP users without historical VoIP usage in 3 months to offer relevant bundles and prevent IDD cannibalization

Gurgaon, Haryana, India

· Developed a predictive model to identify prepaid subscribers at risk of becoming inactive to enable corrective actions to prevent churn using Random Forest· Built Behavioral Segmentation of subscribers based on the historical usage and behavior for a leading Telecom Player in Nigeria using k-means clustering technique· Built Predictive Models to identify severity of vulvovaginal atrophy in women using ensemble learning

Gurgaon, Haryana, India

· Worked with the AbsolutData Labs team to create Classification API for the NAVIK AI web platform of the company · Classification API enables users to perform basic exploratory data analysis like build inference plots, distribution plots and build machine learning models for any dataset having a binomial dependent variable · The user can perform basic data audit and EDA, outlier treatment, missing value imputation, calculate variable importance and build predictive models using multiple techniques like Random Forest, Naïve Bayes, Logistic Regression

Master Of Science - Ms, Business Analytics

Bachelor Of Engineering - Be, Electrical And Electronics Engineering