Quant discretionary long/short absolute return fund powered by data science

Responsible for designing and implementing business solutions leveraging Machine Learning and Quantitative methods using tools like Python, Spark, Hadoop, Hive and SQL.LABOR PRODUCTIVITY MANAGEMENT (Production): Machine Learning application built with PySpark and Python designed to evaluate amount of units handled by warehouse workers in near-realtime based on 60+ variables in warehouse and complexity of tasks.• Designed and coded 60+ features and Machine Learning model solution in Spark• ‘Best model’ picked after multi-variable/feature grid search and 10-fold cross validation• Model designed to train every night and generate predictions every hour• Model reading data from Hive tables on Hadoop• Predictions tested throughout the pilot covering multiple scenarios over two weeksINBOUND ESTIMATION (CARTON PREDICTION): POCMachine Learning Application designed to predict inbound 10 days in advance in order to optimize staffing. Part of supply chain visibility effort.

Use Monte-Carlo simulations and stochastic calculus to model expansion of Supernovae in our Galaxy