Work Experience
Work Experience
I am currently a Research Assistant working with Dr. Yanjun Li and Dr. Matthew D. Disney, focusing on Deep Learning Methods to computationally predict RNA-Small Molecule Binding sites.
Previously, I have worked at startups developing products and conducting research.
My relevant experiences are listed below. You can also take a look at my Resume.
Research Assistant
September 2023 - Present
University of Florida - AI Driven Drug Discovery Lab
Led a bioinformatics project on RNA small molecule binding site prediction, leveraging graph neural networks and transformers to surpass benchmarks on T18 and T9 datasets by 7%.
Built deep learning models that effectively utilized atomic information and attention mechanisms to integrate 3D RNA structure data, achieving best-in-class performance on the novel Hariboss dataset by 21%.
Implemented efficient training techniques including LoRA (Low-Rank Adaptation) for transfer learning, and engineered comprehensive data pipelines for parsing and cleaning complex data from Protein Data Bank.
Developed a Recurrent Neural Network & LSTM model using PyTorch to predict exercises from Pose Estimation data, improving prediction accuracy by 20%.
Engineered mathematical models in Python to calculate exercise repetitions and pacing, achieving 95% accuracy in real-time tracking.
Integrated ChatGPT API and constructed Retrieval-Augmented Generation (RAG) models, enhancing the app's conversational capabilities and information retrieval accuracy by 40%.
Optimized the app's text-to-speech service and implemented custom post-processing algorithms, reducing processing time by 50% while increasing output accuracy by 10%.
Machine Learning Engineer
August 2022 - July 2023
RC3 Canada Inc
Full Stack Developer
January 2022 - July 2022
Developed key features for the company website including a job portal, search functionality, and team page using React, improving user engagement by 40%.
Collaborated in designing and implementing a MySQL backend database, optimizing query performance and reducing data retrieval time by 35%.
Engineered efficient API calls between frontend and backend using NestJS, resulting in a 50% decrease in load times.
Inovatyv
Data Science Intern
January 2022 - July 2022
Developed a Graph Neural Network using PyTorch to classify prescription data into Medicines, Symptoms, and Lab Tests, achieving 96% accuracy.
Engineered custom features based on spatial and temporal analysis of prescription writing patterns, enhancing model performance and enabling accurate predictions across diverse data.
Scaled the solution to accommodate 2,500 doctors across 200 hospitals by implementing four generalized models, reducing manual data entry by over 90%.