Pratik Sonar
About Pratik Sonar
•A visionary technologist with 2+ year of experience in building a cloud-based product which uses highly sophisticated AI/ML-driven Next Best Action Recommendation Engine designed to deliver personalized omnichannel journey. •Built a Product/Toolbox of advanced analytical models, which can empower pharma companies to make intelligent and personalized decisions that seeks minimal intervention from user while designing best fit solutions. •Incorporated Data Mop module in Product with automated way which cleanse and preprocess the data. •Incorporated Customer Persona/Segmentation module in Product with automated way to identify target markets •Incorporated Channel Persona/Channel Affinity score module in Product with automated way which helps in identifying customers with high affinities to a certain channel •Incorporated Best Time and Day module in Product with automated way which helps in identifying best time and day for reaching out each HCP channel wise to increase customer engagement. •Incorporated Market Mix Model in Product with automated way to determine the optimized channel mix for different scenarios •Incorporated Lead score module in Product with automated way which helps in identifying the propensity of writing prescription with the given channel engagement of each customer at individual and segment level. •Developed an Email Performance end-to-end classification system using Random forest classifier for predicting probability whether a marketing email will be opened or not and once opened whether it will be clicked or not based on features extracted from email subject lines, email bodies, delivery times and customer demographics. •Incorporated Engagement score module in Product with automated way which helps in identifying the measure of engagement of each customer at individual and segment level. •Developed Content Analytics module which helps in personalizing/recommending the content/key messages shared via digital communication by understanding the HCP affinity toward different topics for digital assets by finding key message affinity score. •Incorporated Channel sequence module in Product using Markov Chain and recurrent neural network (RNN) LSTM