Ashrith Reddy Email & Phone Number

Boosted the Analytics maturity of Emerging Technologies Product division (Augmented/Virtual Reality) of leading US hypermarket chain.- Architected and renovated the Data pipelines to meticulously capture all Website/App user activity of Augmented Reality to enhance the coverage by 120+ times. Engineered end-to-end flow from the origin of data to business.

Developed and implemented a Bayesian Regression framework, designed as a versatile, plug-and-play solution to seamlessly integrate with an expansive in-house ML ecosystem of products.– Engineered the model to be highly adaptable, catering to a diverse range of industry-specific use cases, while ensuring compatibility and coherence with the existing suite.

• State Transition for Customer Analytics team of leading US Hypermarket Chain
• Developed a Markov-chain simulator to assess the impact of promotion (which alters the customers' distribution during acquisition) on revenue and customer count. The Markov chain state here is defined as count of.
• This simulator supports the business in decision making on choice of promotions to offer while acquiring customers and when.

• Liaised a leading Education-Tech company that helps US universities identify and enroll students for graduate medical programs.
• Applicant Fit and Yield: Identified the drivers of an applicant's admission and matriculation. Implemented multiple statistical techniques (like Logistic Regression, Random Forest and SVM) to determine if an applicant.
• Applicant Churn: Real-time identification of applicants who would attrite and nudge them to complete the application. Model steps involved variable importance using Random Forest, recursive feature elimination.
• Applicant Strengths and Weaknesses: Applicants typically aren't aware of their areas of strengths and improvements when applying to graduate programs. Logistic regression models were built to identify attributes of.

• Price Elasticity and Optimization for Value Fashion Retailer
• A leading retailer wanted to determine price elasticity of their products to understand price-demand relationship and drivers of sales, and to further calculate optimal products' price and maximize margin.
• This analysis involved determining a suitable level of analysis, outlier treatment and other exclusions, Product Lifecycle identification, selection of Model technique and transformation, Model building & validation.
• Experimented with various regression techniques (like OLS, Lasso and Ridge) and ElasticNet with log transformation was chosen to model for units sold.
• Recommended prices with various confidence intervals using bootstrapping, as ElasticNet being a penalized technique does not yield p-value.

• Promotion affinity and Customer Segmentation for tween Apparel retailer
• A major US apparel retailer wanted to understand how their customers perceive different price promotions.
• Customer Segmentation: Used K-means clustering technique in Python to segment customer using their product affinity and purchase behavior patterns.
• Promotion Analysis: Calculated response rate and revenue lift to determine customer segments’ affinity towards each of the promotions.
• This enabled them to understand the behavioral traits of customers towards promotions and make informed price promotion decisions based on customer segments.

Built scalable & automated modules in R and Python for Univariate analysis, Bivariate analysis, conversion of hierarchical data to tabular format, data extraction from leading 8 social media sites, APIs and open sources, optical character recognition, etc.Designed coursework and taught Advanced R-programming and Python for data analysis to 100 trainees.

• Worked with a major US sports broadcaster to enable serving customized advertisements based on sports viewers’ emotions, in addition to achieving better ROI for the advertisers. This analysis helped increase.
• Prioritization of various events in sports: Identified major events (e.g. Touchdown, Grand Slam etc.) occurring in big-6 American sports leagues that elevate emotional state in viewers and ranked them by the impact..
• Behavioral Customer Segmentation: Designed an accurate RFM methodology to identify a viewer’s team and league preference to complement the existing method of geographic location and favorites. It involved analyzing.
• Analysis of betting odds: Extracted odds data for all the games to identify spread or money line thresholds to classify a future game as ‘Tight Game’, ‘Expected Win’ and ‘Expected Blowout’. Variables considered: Spread.

• Gains and Losses analysis for largest US Hypermarket chain
• A leading US retailer has seen a decline in cold cereal category sales for the past few years. Analyzed transaction data for cold cereal and other food categories to identify the causes of declining sales and.

Master Of Science - Ms, Statistics, 3.7/4

Activities and Societies: Graduate Assistant (Department of Statistics): Led the development of a plug-and-play academic program website.

Bachelor Of Technology (B.Tech.), Electrical And Electronics Engineering

Activities and Societies: Engineer (Technical Fest), Incident (Cultural Fest)Computer Programming and Lab (C++), Calculus I, Calculus II.