AI/ML Platform

Analyst – Flipkart Search (Ranking and Semantics)• Sole analyst looking after Flipkart’s Search RankingProjects• Personalization: Launched Personalization in search results based on user segments• Sale-specific Ranking: Designed a mechanism to show products with good sale-potential higher in the search results during sale events• Diversity: Diversified products in the search results based on product attributes• Optimization: - Cold-start: Proposed changes to the existing ranking algorithm to improve the discovery of cold-start products (products with less views) in Search Ranking results - Discovery Method Normalization: Proposed a method to normalize the search ranking algorithm for the different methods of a product's discovery on Flipkart based on user intent - Profitability: Analyzed the different discovery methods to identify the opportunities for profitability• User’s Journey: Performed an analysis on a user’s journey when he/she searches for a product and how can Search improve based on the user’s behaviour• Returns: Identified the various reasons and proposed solutions to customers returning a product on 2Gud even before the product is delivered using Random Forest and Logistic RegressionTools/ Technologies: Python, Adobe Analytics (Omniture), PySpark, Hive, Excel, R

Software development for Point of Sale (PoS) terminals• Designed and developed a PoS application in C++ from scratch which was later deployed to PoS terminals• Wrote code to handle communication between PoS terminals and an FTP server over SSL• Designed packets and implemented them for Pin Pad devices which enabled interaction and data transmission between PoS terminals and Pin Pad devices• Handled end-to-end ownership of modules’ design and development• Handled setting up, configuring and maintaining an FTPS server• Triaged issues raised by customers• Automated various processes in Python• Conducted Knowledge Sharing sessions for team membersLanguages: C++/CPlatform: UNIXVersion Control System: Git