Red Hat, Inc.

Business Data Scientist - 2 • June 2021 - Present

• M&A Timeline Prediction:
Instituted ML driven approach for startup acquisition timeline prediction using market research data

• Marketing Intelligence & Country Prioritization:
Primary focus of the project was to understand where should Red Hat invest its next incremental Dollar($) The marketing intelligence data set used for the project was a repository of internal and external metrics related to Red Hat businesses such as addressable market data, macro and micro economic data as well as internal data on a per country/language basis. The goal of the project was to see if the existing way of ranking countries and languages can be optimized & enhanced using traditional or nontraditional statistical methods (regressions, time series, machine learning, AI). The ranking consisted of ranking countries and languages based on the probability to be successful for Red Hat business according to past performances in those countries/languages. ML Techniques Used: Regression Analysis, Random Forest (Weight Estimation of features), L2-Regularization

• Long Term Revenue Forecast:
Forecasting business revenue based on internal and external metrics is a key factor to develop coherent and data driven strategy and decision at Red Hat. This project was used for analysis of different kinds of forecasting methods using internal(marketing, financial & sales factors) and external metrics(Macro economic factors) for analyzing which methods are more appropriate depending on the products, the region or other criteria. Leveraging advanced Time series & ML techniques like ARIMA, LSTM Recurrent Neural Networks and feature engineering techniques like PCA was the key part of the project.

Amdocs

Datawarehouse-Business Intelligence Developer • July 2016 - July 2019

Major Tools : Teradata, HP Vertica, Oracle DB, Informatica, Unix/Linux Shell Scripting
AT&T Sales PCA Data Mart:
This application was maintained for keeping sales data of AT&T customers where my major work contribution was towards redesigning and developing the data-mart to get the Near Real Time(NRT) Sales data for generating better insights on new AT&T customers and further performing customer segmentation using these insights.

Major Tools : Data warehouse Teradata, HP Vertica, Informatica, Unix/Linux Shell Scripting
• Primarily, worked with AT&T Marketing Team on Business Intelligence solutions • Developed processes and evaluated data of over 1+ million users for identification of actionable marketing insights in telecom billing, viewership, sales and product-feature data of AT&T customers • Improved data gathering, data munging & analysis process for CRM applications • Developed CRM data pipeline for AT&T's TV Now product integration(2017)
AT&T Integrated Marketing Data Mart:
This application was created for identifying actionable marketing insights using demographics data, product-feature data, event viewership data, billing data & product subscription data for DTV, Uverse, Mobility & TV Now products. 1) Primary work contribution was towards building data infrastructure for DTV and RC1 conversions and later for integrating AT&T TV Now product data into the application. 2) Churn-Prediction of AT&T customers based on various market segments 3) Building the data infrastructure from marketing perspective during Time-Warner Merger for understanding prospective customers