Thang Tran

Business Data Analyst

Thang Tran headshot

Turning messy data into clear decisions with SQL, Python (pandas), and BI (Power BI / Tableau).

About Me

Welcome to my website! I’m Thang Tran, a UMass Lowell graduate with a degree in management information system. I am passionate about turning complex data into clear, actionable insights and making web pages. Through my co-op at Lightforce Orthodontics and multiple analytics projects, I’ve gained hands-on experience with SQL, Python, Power BI and Tableau. I hope we can meet soon!

Get in Touch

Projects

Selected work with outcomes. Ask for code samples or deep-dive writeups.

Sales & Profits Analytics Dashboard

Python • Excel • Power BI

Sales & Profits Analytics Dashboard screenshot

Developed to explore sales, costs and profits across cities, products, managers and purchase types.Includes dynamic filters, map visuals, KPI cards and interactable elements so that stake holders can understand the data, uncover trends and quickly identify profit drivers. Cleaned and prepared raw excel data and used DAX measures to calculate interactive KPIs. This dashboard would essentially allow restaurant managers to monitor real-time performance across different locations, track profitability trends and identify which products were high performing

Holiday Impact on Retail Sales

Python • Excel • Power BI • SQL

Holiday Impact on Retail Sales screenshot

Enabled managers to monitor $21.7B sales as data across 45 stores, improving visibility into profitiability trends for holiday season. Authored an end to end retail analytics pipeline with SQL Server, Processed and transformed 6,000+ rows with SQL bulk inserts and star schema modeling, type safe SQL in order to designed and deliver a clean interactive Power BI model with multiple pages showing analytics on real world data such as growth rate, total sales over time.

Netflix Data Mining Workflows Project

Rapidminer • Microsoft Excel • Python

Netflix Data Mining Workflows Project screenshot

Analyzed and cleaned nearly 9000 Netflix shows and movives using the Naives Bayes model to uncover patterns in genre, country of origin and ratings with a accuracy of 99.97%. Tested decision tree, logistic regression models as well, but they did worse than Naive Bayes.Found movies dominate over TV shows; dramas, comedies and documentaries are most common; US, India and UK lead in production.

Experience

Roles where I delivered measurable impact.

Contact

Want to talk about an opportunity or a project? Reach out! I’m based in the Greater Boston area and open to roles.

Send a Message