GenAI Capabilities to Query Millions of Unstructured Documents for Research and Clinical Care.

· Architected and launched the company’s first production RAG system for semantic retrieval across unstructured clinical data.

· Designed and deployed containerized data pipelines for efficient document chunking, embedding generation, and vector indexing.

· Built and optimized a vector search infrastructure on GCP.

· Established data quality monitoring and observability using Elementary and DBT.

· Developed production-grade APIs and testing frameworks supporting event-driven agentic workflows and data platform services.

AI-powered Clinical Trial Matching for Precision Oncology.

Transformer-based NLP Models Deployment and Serving, Data Orchestration and Modeling via Airflow/DBT, Clinical Data De-Identification, Batch Incremental Data Ingestion Pipeline, Docker, GCP, Kubernetes, etc.

ML modeling(classification, survival analysis, time series prediction, patient engagement, retention), ML deployment, Data(ETL)/ML pipeline, REST API(Flask), EHR/Longitudinal data, AWS/GCP, Clinical risk, Claims analysis.

I built CareSignal's first ML solution from the ground up. CareSignal AI is an ML solution with predictive models for chronic disease management in a remote patient monitoring setting. Also, I keep working on maintaining/improving current models, developing new models, and deploying models into production.

I leveraged Machine Learning algorithms to find cognitive patterns for pediatric brain tumor patients. I did my research in the NeuroMedical Informatics Lab by Dennis L Barbour, MD, PhD.

Specifically, I: