At DramWise, I built ETL pipelines that automate financial data processing across five bank statement formats. I also designed a PostgreSQL schema to improve query performance for historical analysis.
As a freelance Machine Learning Engineer, I developed end-to-end ML pipelines from exploratory data analysis through model evaluation, with Power BI dashboards for model monitoring and stakeholder insights. Earlier, as an Intern Statistical Programmer at Gurus LLC, I standardized and preprocessed clinical trial data in R to meet compliance requirements.
On my ChEMBL Molecule Similarity Pipeline project, I built an Airflow-orchestrated ETL pipeline on AWS to compute molecular similarity across 2.9M+ compounds. My data science capstone examined how image color formatting affects deepfake detection accuracy, finding that YCbCr improved accuracy for certain manipulation types.
