These projects show my current work at the intersection of data leadership, AI systems, analytics products, and practical decision tools. Some are live applications you can use directly in the browser; others are technical writeups and agent experiments.
I focus on projects with a clear business question, a working product, and a technical system underneath it that can be explained, extended, and improved.
Current Projects
2026 Soccer World Cup Predictor - Featuring GenAI
Live Application
A working simulator for exploring 2026 World Cup scenarios, built as a live AI-assisted sports analytics product.
Live Streamlit App
Streamlit Community Cloud blocks reliable iframe embedding on this site, so the app opens in a dedicated browser tab.
2026 NBA Finals Predictor - Featuring GenAI
Seasonal Application
An interactive Streamlit app for simulating NBA Finals outcomes with a GenAI-assisted analysis layer, preserved as a completed project now that the Finals are over.
Completed Streamlit App
The app may be asleep on Streamlit Community Cloud, but the project writeup documents the build and the AI-assisted workflow.
Recent AI Writing
Published in Data Science Collective
Current AI Systems & Agents
Active Work
- Bobby G Agent - Personal Data Scientist + Operator
- MCP Project
- Hermes AI News Summary Agent, built using ChatGPT
Project Archive
Keepa API: Amazon Price Data
Problem Summary:
Amazon sellers need access to real-time analytics
My Approach:
I build a custom tool using the Keepa API to dynamically deliver analytics on a given Amazon product.
Result:
Link to my project writeup is below
Buy the Amazon Price Data - Keepa API
How I Built My Own Fitness Tracker Using Google Fit Data
Problem Summary:
Fitness data is stored in Google Fit but not easily visualized anywhere else
My Approach:
I build a custom tool using Google Sheets to visualize the data
Result:
Link to my my Towards Data Science writeup: "How I Built My Own Fitness Tracker Using Google Fit Data"
Project Using a Data Pipeline in Google Apps Script
Read the Article
Predicting Stock Price Using Natural Language Processing
Problem Summary:
Each quarter CFOs and CEOs of major corporations talk to investors via an earnings call. This text data can be used for predicting stock price.
My Approach:
I used NLP libraries like TextBlob and ELI5 to do sentiment analysis on text data
Result:
A Comprehensive ML experiment that uses sentiment analysis to predict stock price
Link to Medium Article