- data pipelines for RAG, FT, and LLM evaluation- fine tuning transformer models with industry partners

• Developed tool that identifies trending eCommerce products, and allows wholesale companies to identify, design, and manufacture similar products • Acquired and cleaned data from social media and eCommerce websites• Developed model using neural networks to match similar products across the web

• Developed new method combining multiple linear models with reduced dimensionality that accurately simulates protein folding, with an order of magnitude speedup in simulation efficiency.• Using k-nearest neighbors, identified key differences in lipid packing defects in bilayer and monolayer membranes• Using these models, identified key mechanism for preferential targeting of proteins to lipid droplets using molecular dynamics simulations, a key step in metabolic… Show more • Developed new method combining multiple linear models with reduced dimensionality that accurately simulates protein folding, with an order of magnitude speedup in simulation efficiency.• Using k-nearest neighbors, identified key differences in lipid packing defects in bilayer and monolayer membranes• Using these models, identified key mechanism for preferential targeting of proteins to lipid droplets using molecular dynamics simulations, a key step in metabolic processes.• Collaborated with research group at Harvard Medical School to validate computational predictions. Show less

• Developed highly detailed additive model for protein mimetic molecules• Investigated folding of protein mimetic molecules in order to identify key aspects of biological proteins. • Awarded 2 competitive research awards.

• Developed in-house customer relationship management (CRM) mobile and tablet application for financial sector customers using Objective-C.