About

Bhumika Singhal is a highly skilled Senior Data Scientist with over 4.62 years of relevant experience in the field. She holds a Bachelor of Technology degree in Chemical Engineering from the Indian Institute of Technology, Guwahati. Bhumika is an expert in Machine Learning, Python, NLP, Computer Vision, AWS, Spark, Tensorflow, Reinforcement Learning, Random Forest, Deep Learning, Pytorch, Apache Airflow, AI, Docker, Keras, Adversarial Networks, Django, Neural Networks, SQL, Search, GAN, Scikit, Big Data, and Java. Currently, Bhumika is working as a Senior Data Scientist at Bright Money, where she is building systems that make intelligent decisions to fulfill people’s financial goals aiming to optimize for long-term retention. Her work involves optimizing goals, budgeting, smart payments, user profiling, overdraft mitigation, unauth prediction, fraud prediction, user churn, and UX modeling. She has designed and built collective intelligence and peer learning models to increase users’ surplus based on their cashflows, spending patterns, and peer data. Bhumika has also designed, modeled, and productionized machine learning-based Underwriting Policy for Bright’s debt refinancing product. Additionally, she has built various models like segmentation models for loss forecasting to predict delinquencies and project losses. Prior to her current role, Bhumika worked as a Data Scientist at Bright Money, where she designed and productionized Debt Manager, an E2E AI system that aims to pay off the liabilities of users by analyzing and modeling their cashflows and saves money in a distributed fashion when affordable. She also built various zero-to-one products like Income/Expense Prediction, Balance Forecasting, affordability prediction, overdraft prediction, collection models, and user churn models. Before joining Bright Money, Bhumika worked as an Analytic Science Scientist II at FICO, where she was responsible for expanding and optimizing data models, prediction algorithms correlation algorithms, as well as text analytics models. She also researched and implemented the necessary frameworks and end-to-end pipelines to generate synthetic data for different types of transaction data using Generative Adversarial Networks. Bhumika developed an end-to-end pipeline for Financial Crimes Studio to educate external data scientists to build real-time machine learning models using FICO’s patented streaming analytics technology. Bhumika is a problem solver and an experienced ML Engineer and Data Scientist. She is a quick learner and has a passion for exploring new