Timeline

UC San Diego Health (The Nemati Lab) [Apr 2023 - Jan 2024]

Collected vitals and fitness data from wearables such as Fitbit and Apple Watch using Fitbit API and Google Fit API respectively.
Implemented Speech-to-Text and Text-to-Speech engines to build a voice assistant prototype that asks healthcare questions and listens for answers.
Developed sophisticated DL models to predict sepsis using the wearable device data.
Utilized grounding LLMs to understand patient answers parsed by the voice assistant and provide additional information to the DL models.
Spearheaded research on patient readmission prediction using multi-modal data including EHR data, wearable Fitbit data, and novel Multi-scale entropy (MSE) features.
Translated MATLAB codebase to Python for internal purposes of the multi-scale network representation (MSNR) project.

CSE 256 - Statistical Natural Language Processing, Prof. Nakashole
CSE 291 F - Topics in Computer Science and Engineering - Unsupervised Learning, Prof. Dasgupta
CSE 293 - Special Project in Computer Science & Engineering, Prof. Nemati

CSE 250A - Probabilistic Reasoning and Learning, Prof. Berg-Kirkpatrick
CSE 258R - Recommender System and Web Mining, Prof. McAuley
MED 277 - Introduction to Biomedical Natural Language Processing, Prof. Hogarth & Prof. Nemati

Guided students through basic Python and ML assignments, and project on predicting on-time performance of flights.
Worked on revamping multiple pages for their website.
Researched on disease detection problem statements including Bradycardia Prediction and Atrial Fibrillation Detection

Took up assignments on Python and ML Basics
Worked on Flight-Delay Prediction project that employs a two-staged pipeline architecture to classify and predict the delay of flights in minutes.