• Implemented a microservice to automatically create, segment, clean and organize data from multiple client facing sources and collate into an organized database structure.
• Wrote backend scripts to pipeline the database data using Apache Spark into machine learning prediction models for different types of sources directly affecting client facing services.
• Formulated prediction models for different metrics such as Ecommerce sales, cashflow, etc. using XGBoost, skLearn and RandomForestRegressor libraries.
• Used Docker containers to deploy the prediction models onto real-time client services that improved user experience and retention inside the product.
