Roshan Shetty

Provide marketing analytics solutions to the online grocery division of one of the world's top retailers. Projects include: - Analyzed marketing campaigns for online grocery with Python, SQL, Hive, Spark, Tableau and Adobe Analytics; better visibility and improved tracking helped in getting 160% weekly lift in revenue on ad spend (ROAS) in one of the recent campaigns this year - Built lists of subscribers facing issues due to COVID demand surge in very tight timelines (turnaround time < 24 hours); lists were targeted and post analysis showed 7% lift in conversion from trial to confirmed compared to control - Built forecasting models to predict how retention rates will trend over time using Prophet library in Python - Analyzed effectiveness of direct mailers using an a/b test setup and shared insights with the business - Built retention metrics to understand how marketing is performing in bringing back lapsed and churned customers

Built machine learning based response and approval models to optimize direct mail campaigns in Python

• Designed and co-implemented a follow up score on current leads which allow the sales development teams figure out which leads need to be worked with at the current moment • Analyzing product usage of our current customers to see how much value we are providing to our current customer base to allow opportunities for up-sell

• Came up with lists of accounts that fit the company’s ideal customer profile from various sources like Crunchbase, DiscoverOrg and BuitWith • Designed and managed the implementation of a health score of our current customers based on their usage of our product • Built lead saturation reports that helped gauge whether a lead had a quick and a high follow up thereby allowing MDR teams change strategy accordingly

• Built lead scoring models using Python which will predict the probabilities of conversion and closure of the leads • Worked with developers to increase the amount of data tracked on the company website from 20 to 80% • Built a first touch attribution model using SQL and Python which increased lead count from 650 to 1300 in 3 months

As a Udacity Project Reviewer, I review student project submissions, striving to give actionable and helpful feedback.

• Building predictive models to predict session registrations in various sub-tracks at SAPPHIRE 2016 conference using R • Building propensity models to cater to marketing line of business using SAP Predictive Analytics • Built a logistic regression model to predict when the business warehouse platform will respond slowly thereby improving the system performance by prospectively fixing issues • Building forecasting models to predict end of quarter sales revenue of SAP North America region to within 10% error

• Implementing a prototype of a resource allocation and leave planning mobile dashboard with potential to provide significant cost savings and increased efficiency • Currently working to deliver key enhancements both on-premise and in the cloud, that would enable the Opportunity to Quote process to work more effectively

• Working with a team of up to 12 newly-enrolled Udacity Data Analyst Nanodegree students to develop positive group dynamics and set expectations for weekly progress • Facilitate group problem solving and friendly connection leading to strengthening students’ concepts and making students independent learners

• Initiated trainings in SAP UI5 and led the development team in the implementation of a system which executes retail promotions based on the expected return of investment for the Senior Management • Developed a new UI using HTML, CSS and SAP ABAP quickly and efficiently for the Senior Management which was used in 5 client pursuits out of which 3 turned out to be wins worth over 5 million dollars • Implemented an interaction center portal using SAP CRM for one of the world’s leading household products manufacturer which enabled it to reduce the customer service interaction time by 24% • Upgraded the SRM portal of a leading pharmaceutical company using SAP Web Dynpro ABAP which improved the performance of the system by 32% and thereby quickened the procurement process • Extracted, transformed and loaded (ETL) data for the world’s leading propane supplier using Excel which eliminated the errors in data

A set of 6 projects have been implemented covering search, planning, HMM's, CNN's, RNN's and NLP

A set of 7 projects has being implemented which covers a broad range of technologies used for Data Analysis and Data Science.

• Recipient of Graduate Student Scholarship (GRE: 327/340) • Successfully completed the ‘Data Analytics’ certificate • Represented UC to a top 6 finish from over 20000 teams in the Boston regional finals of Hult Prize 2015

• With interests in Signal Processing, I pursued my mini project in a Compression Sampling System with end-to-end simulation and partial implementation using hardware. • Participated in various robotics competitions and workshops in college. Applied various concepts of electronics in building automated robots.