• Leading a cross-functional team of 25+ engineers in the Core Application track of the product. Directlyresponsible for planning, execution and release of projects on the roadmap. - Delivered multiple projects to integrate an acquired entity into CleverTap. Expected to capture a large majority of the acquisition’s revenue and move it to CleverTap’s accounts.• Mentored 15+ new engineers in the company who have grown to lead and work in 4 fully functional teams.• Identified and implemented changes to the development process, thereby reducing QA efforts by 20%. Changes rolled out and adopted by all teams in the organisation.

• Built a cohesive team of 10+ engineers from scratch. Acted as the team lead, where I led the planning and execution of projects for the "Journeys" sub-product. - A content generation tool designed for marketers to reduce copy generation time from days to minutes. - Enhancements to increase push notifications render rate by 15-20 percentage points, thus uplifting it to 85%. - An experimentation framework which enables clients to design customer journeys optimised for conversions.• Collaborated with product, design and QA teams with an active involvement in defining project scope and solutions.

• Built various features in close collaboration with product and design, with a focus on stability and scalability. - Integration with Xiaomi, Huawei and Baidu push notification services to improve message delivery by∼10 percentage points. Used to dispatch 1B+ notifications everyday for 350+ clients. - Integration with WhatsApp partners like Gupshup to deliver 5M+ messages for 350+ clients everyday. - Revamped the journey canvas used to design messaging flows for end users. Leveraged by 1000+ customers.• Selected from 300+ employees for CleverTap’s STAR Awards for valuable contributions to the product.

Project: Supervised Response Models for International Burn Markets• Modeled the response of card members on various redemption options in order to personalise membership rewards. Beat the existing capability by 7.6 percentage points, with work done implemented in production.• Designed an end-to-end solution - from processing raw data to modeling and recommendation.• Engineered features to capture user preferences using redemption, browsing and spend history.

Project: Transport Mode Detection• Processed raw GPS trajectories to design a transport mode detection model from scratch.• Partitioned each journey into uni-modal segments and predicted the mode of transport for each segment.• Extracted sophisticated features, depending on characteristics of a vehicle, to differentiate between 6 modes of transport. Achieved a test set accuracy of 88.3%.• Used grid search and cross-validation to tune relevant parameters and select the best model.

Project: Churn Prediction using Social Network Analysis (telecom domain)• Analysed call records to identify social groups and predict customer attrition. Increased model coverage from 73% to 81%.• Modeled the social network as a graph and used it to extract features affecting churn propensity.• Work done incorporated into the production model and used to predict attrition from over 90 million subscribers.