Tom J.

Developed a Model Serving Pipeline that can integrate multiple models to serve multiple prediction requests. Python expert with AWS experience in deploying microservice applications. Build a text classifier using LSTM, for assigning the consumer tickets to the right department. Authored the application’s Dockerfile and docker-compose to build and deploy. Applied PySpark and Hive-QL for automating rule validations, and Sqoop for ingestion.

Worked on Risk Assessment for an Insurance client, where target label was Imbalanced. Created Classification Models (Logistic Regression, Random Forest Classifier, XGBoost) and evaluated their performance. Structured and authored end to end automated data pipeline on Big Data Platform, including ingestion, cleaning/pre-processing and model building. Associated with Data Scientists for Creating Plots, Feature Analysis, Outlier Analysis, Hyper Parameter tuning and Model Evaluation