About Me

Hi there! 👋 I'm Steven Tran.

My professional headshot (from before I graduated college 😲)

I'm a young, enthusiastic data professional living and working in sunny Phoenix, Arizona 🌞. I have approximately 5 years of professional data analytics exerience, with 3 of those being in the realm of predictive analytics and machine learning in the taxation domain.

Professional Experience

Senior Analytic Consultant

Voyatek 01/22 - Present
  • Used SQL, Python, Hive and Spark to engineer data pipelines, create analytic datasets, and implement and evaluate machine learning models.
  • Led clients through the Cross-Industry Standard Process for Data Mining (CRISP-DM) lifecycle.
  • Trained client and internal modelers to use adopted tools and methodologies.
  • Wrote documentation for processes and procedures.

Economist

Arizona Department of Revenue, Office of Economic Research & Analysis 03/18 - 07/21
  • Automated data entry and validation using Excel, Access, and VBA, reducing process times by up to 80%.
  • Developed JavaScript and VBA scripts to scrape legislative data, and used T-SQL and SSRS to build business reports and dashboards.

Auditor

Office of the Auditor General of Arizona, Division of School Audits 05/16 - 03/18
  • Analyzed financial and performance data of public school districts using Excel and Access.
  • Conducted inspections and interviews resulting in comprehensive audit findings.

Projects

Here are some of my recent projects!

Identifying Nutrient Deficiencies in Satellite Farmland Imagery

  • Trained a U-Net Convolutional Neural Network on 1,300 aerial crop images to classify nutrient deficiencies.
  • Implemented a custom loss function for optimized model performance.
  • Utilized Python, NumPy, Pandas, Matplotlib, TensorFlow, and Keras.

Reddit Post Classification

  • Collected over 10,000 Reddit posts using the Pushshift API.
  • Achieved 80% classification accuracy using term-frequency-inverse-document-frequency vectorization and various classification models.
  • Utilized Python, NumPy, Pandas, ScikitLearn, Matplotlib, Seaborn, and NLTK.

Ames, Iowa Housing Price Prediction

  • Used multivariate linear regression to predict housing prices with an average error of $25k.
  • Utilized Python, NumPy, Pandas, ScikitLearn, Matplotlib, and Seaborn.

Education

  • Economics, Bachelor of Science | Arizona State University, Tempe Arizona
  • Communication, Bachelor of Science | Arizona State University, Tempe Arizona

Recent Blog Posts

Check out my latest (random) thoughts and musings. [All blog posts]

Reflections on my progress so far

November 17, 2024

Reflections on my progress so far