Thomas Rivera

Architected and deployed robust machine learning pipelines to support CRISPR gene-editing initiatives across multiple therapeutic programs. Built and optimized NLP models using BioBERT, Hugging Face Transformers, and spaCy to extract gene-disease relationships from biomedical research papers, patents, and clinical trial documents.

Led the integration of ETL processes with MLOps pipelines, focusing on the deployment of AI/ML models. Developed and fine-tuned GPT-3.5, GPT-4, LLaMA, and BERT models to enhance customer engagement and optimize predictive performance for tasks like customer churn analysis and sentiment classification.

Engineered scalable fraud detection models for card transaction analysis using autoencoders, Isolation Forests, and LSTM for time series pattern modeling. Processed 50M+ transactions daily via Apache Spark and Hive, reducing fraud detection time from 5 minutes to under 1 second.

Built ML models to predict patient readmission risk using logistic regression, Random Forest, and survival models, reducing 30-day readmission rate by 12%. Merged patient vitals, EHR notes, and socio-demographic data to construct holistic risk profiles for over 80,000 patients.

Obtained a Bachelor of Science in Information Technology, building a strong foundation in IT principles and practices. Developed skills essential for a career in technology and data management.

Completed a Master of Science in Computer Science, focusing on advanced topics and research within the field. Gained expertise in various computational theories and practical applications.