Jordon Troxell
I’m Jordon Troxell, a data analytics and business professional with a strong foundation in market research, performance analysis, and data visualization to support strategic decision-making.
I hold a B.S. in Marketing with a Data Analytics specialization from DeSales University and am currently pursuing an M.S. in Data Science at Indiana University.
My work focuses on translating complex data into clear, actionable insights that align with business objectives and drive growth, efficiency, and innovation across marketing, product, and operational environments.
Projects
A/B Testing for Marketing Campaign
Analyzed the performance of two marketing campaigns using A/B testing methods using Python. Data from a control group and a test group were compared to determine which campaign achieved a higher conversion rate and better cost efficiency. The datasets were cleaned, merged, and summarized to calculate total spend, impressions, clicks, and purchases for each group. Statistical tests were then performed to evaluate whether differences in conversion rates were significant. The results showed that the test campaign produced a lower conversion rate and fewer purchases per dollar spent than the control campaign, indicating that the control version was more effective overall.
See the in-depth explanation below for more info!
Live Flood Monitoring Dashboard via U.S. Geological Survey API
Built a real-time flood-monitoring dashboard for the Lehigh Valley that pulls live river-gage data from the USGS API, transforms it in Power BI, and displays current water levels on an interactive map and individual gauge visuals for each location. The dashboard highlights potential flood-risk conditions and provides a clear, centralized view of river activity, demonstrating real-time data integration, API usage, geospatial mapping, and effective data visualization in a practical, community-focused project.
See the in-depth explanation below for more info!
AI Image Classification Web App
This project is an AI-powered image classification web application built using Streamlit, TensorFlow, and a pretrained MobileNetV2 deep-learning model. The app allows users to upload any JPG or PNG image and automatically analyzes its contents using ImageNet-trained computer vision. Once an image is uploaded, the system preprocesses it, feeds it through MobileNetV2, and returns the top three predicted object labels with confidence scores. The interface is clean, fast, and interactive thanks to Streamlit’s UI components, caching, and live execution model. Overall, the project demonstrates practical end-to-end machine learning deployment — from loading a pretrained neural network and handling real-time inference, to building a user-friendly web interface that makes advanced deep-learning accessible to anyone.
See the in-depth explanation below for more info!
Let’s Connect
www.linkedin.com/in/jtrox
jordontroxell1@gmail.com
