Arjit Upadhyay

Developed and deployed machine learning models for revenue forecasting, category analysis, and other tasks. Proficient in time series forecasting (SARIMAX, ARIMAX, LSTMs), data stationarity analysis, and NLP tasks. Experienced in deploying models on Azure platforms (transcription, translation) and utilizing Azure Databricks for model training and predictions. Developed a highly accurate (97.8%) time series model for collection prediction, exceeding client's 95% accuracy requirement. Performed comprehensive data exploration and collaborated with clients to ensure data quality. Conducted in-depth data exploration (EDA) for collection and policyholder data. Built and compared baseline models to select the optimal solution for collection prediction. Ensured data quality through collaboration with clients to address inconsistencies and missing data points. Utilized PySpark for data preparation for time series modeling with SARIMAX. Analyzed data linearity using moving averages concepts.