AI Projects
This project aims to develop a user-friendly LLM application with, that answers your questions about legal documents. Simply upload a document (lease agreement, contract form, user manual, rental agreements, privacy statements, etc) and ask your legality questions.
This app, context-aware and history-bearing, will give you relevant answers, help you skip or reduce long reading time so you can lower cognitive load!
Image Caption Generator
AI Legal Document Assistant
This is a captions model for social media posting. Simply upload an image and this app will generate social media worthy captions! The backbone is a RESNET 50 + LSTM combined model that uses both image features and textual embeddings to generate descriptions. Then, the descriptions will be passed to a LLM model for the caption!
This application shows the possibility of combining inputs and outputs from both deep learning models and LLMs in synergy, paving way for more hybrid modeling approaches.
Data Science Projects
Find my previous data analytics & machine learning applications across various domains.
To see project details, click
EV Demand In USA
Web-scraped Data from an enterprise EV models database, and combined multiple public datasets into a single dataset for consumers and policy makers.
Customer Churn Classification
Predicted Risk Scores of Customers who might unsubscribe from services with an ML model so the company knows who to target in marketing efforts.
Time Series
Predicting Air Traffic Demand for the San Francisco International Airport in 2024 with time series modeling and made extensive data visualizations.
Computer Vision (object counting)
Counted agricultural projects in video frames by fine-tuning TensorFlow Object Detection API, writing a video splitting algorithm, drastically reducing tedious human labor.
Computer Vision (object detection)
Trained a customized computer vision model to identify cyclists in images by setting up data transformation, training, and inference pipelines.
Business Analytics for Davis Businesses
Web-scrapped data from Yelp to quantify and understand the businesses and consumer preferences in the City of Davis.
Stock Clustering Project
Trained a customized computer vision model to identify cyclists in images by setting up data transformation, training, and inference pipelines.
Made a comprehensive analysis of Covid-19's impact on unemployment in California Counties. Analysis examined the role of race, gender, and vaccination rates over time across California.
