Things I've built, researched, and shipped
Used Cars Price Prediction for the Vietnamese Market
Investigated and extracted used car listings from Vietnamese e-commerce platforms. Built a machine learning model to help Vietnamese consumers make informed purchasing decisions. The research was published at a national scientific conference and archived with ISBN.
Car Specifications Dataset
Open-source dataset covering 44,934 car models and variations mass-produced from 1985 to early 2022. Includes a Scrapy-based crawler with clear instructions for re-crawling.
View RepositoryE-commerce Product Classification
Classification module for e-commerce product names into four categories. Uses sBERT and phoBERT transformer embeddings with a custom two-layer neural network. ONNX-accelerated inference, deployed on Streamlit Cloud.
AML Data Serving Pipeline
Bank-wide Anti-Money Laundering data serving from datalake at Techcombank. Fabricated 10+ ETL jobs and contributed 200+ features to the Risk Datamart. Part of the Credit Application Fraud detection system.