At Amazon, I built an AI cloud-hosted JIRA ticket-triaging agent that categorized more than 50 weekly tickets across nine teams and 12 components with 92% accuracy. I added it as a Slack bot and developed an AWS pipeline that supported 25+ concurrent requests while reducing triage time by more than 85%.
At BNSF Railway, I created an AI agent that generated SQL queries from natural-language input using OpenAI’s GPT-4o, and developed an OCR pipeline to extract data from PDF reports. At the University of North Texas, I developed an XGBoost model for Nasdaq-100 futures signals, achieving 60% accuracy and 70% precision on the four-hour timeframe.
My projects include Able Faith Mobile, an accessible app with voice navigation and online inventory checkout, and Axworld, a map application that uses Llama 3 to personalize recipes. I’ve also built a Streamlit app to compare phishing-detection models, with the XGBoost model achieving 96% accuracy.
