DitecT Lab at Columbia University

- Building website and databases for Maiv Health using HTML and DJango- Developing secure on-device data storage process for historical measurements database, enabling tracking meaningful trends of vitals and biomarkers- Establishing architecture for security and robust data management (using Django) to support future patients and facilitate data sharing with providers

- Completed self-directed learning to grasp cyber security concepts, project background, and tooling (AFL++ framework, FuzzBench) within 2 weeks- Collaborated with Ph.D. and developed an automated software testing tool, fuzzer, that can detect system security vulnerabilities

- Coded a customizable atmospheric dispersion model (using Python) that can predict the concentration of leaked chemicals based on the Gaussian Plume Equation- Designed an interactive Open-Street-Map centered at the leak source with a heatmap depicting the severity of toxin emission- Deployed the model and visualization to the company website to show the general public about impacted areas in case of any leakage

- Tutored 20+ students through the weekly labs on topics of Absorption Spectrophotometry, and Beer's Law. pH, Molecular Modelling, Qualitative Synthesis, Inorganic Chemistry, etc.- Held weekly session to guide students through their lab analyses, lab reports, and pre-lab assignments for the following week

- Implemented Data Exploration, Data Cleaning, and Data Preprocessing on a dataset from Psychology Foundation of Australia; examples of the above processes are determining and dropping issing data, determining correlations between features, choosing features of interest, and standardizing data;- Implemented linear regression and Convolutional Neural Network to determine the correlation between a potential patient’s mental condition (e.g., depression) and their physical symptoms (e.g.… Show more - Implemented Data Exploration, Data Cleaning, and Data Preprocessing on a dataset from Psychology Foundation of Australia; examples of the above processes are determining and dropping issing data, determining correlations between features, choosing features of interest, and standardizing data;- Implemented linear regression and Convolutional Neural Network to determine the correlation between a potential patient’s mental condition (e.g., depression) and their physical symptoms (e.g., eating disorder) on the data obtained from abovementioned database Show less