Data Scientist - IQVIA
Senior Machine Learning Engineer - Quantiphi Analytics
- Mentored and Spearheaded a team of 7 Machine Learning Interns to build a Proof of Concept.
- Developed a customer attrition model using a large dataset of 20 million records with machine learning algorithms such as XGBoost, Random Forest, CatBoost, and Artificial Neural Networks to predict customer churn based on historical data.
- Implemented and worked with Language Models such as PaLM, T5, BERT, and GPT.
Built an open-source chat application that is capable of conversing with the user as well as documents.
- Focused on ideating, brainstorming, researching, and building innovative AI and ML products, solutions, or workflows.
- Developed a Custom Synthetic Document Generator from scratch that is used to generate synthetic documents that can be used for training ML models.
- Using synthetic data generated by CSDG, I trained a custom processor in Google Cloud Document AI and automated the document tagging process, which reduced the time and manual effort required to find training documents and document annotation by 50%.
- Worked on Document AI services in Google Cloud Platform (GCP) for document understanding, processing, extraction, and analysis.
- Conducted a hypothesis test to prove the correlation between prostate cancer incidence. Employed statistical approaches including z-test and p-value analysis to validate the hypothesis.
- Utilized Random Forest and XGBoost algorithms for predictive modeling to enhance the accuracy of our findings.
- Worked on a proof-of-concept to implement Sentiment Analysis using state-of-the-art Transformer models, including BERT, RoBERTa, DistilBERT, and XLNET