/* Building intelligent applications with machine learning, LLMs, RAG, computer vision, and automation. */
🟢 Open to work
Currently an Extern @WayfairMy portfolio collects work and experiments related to AI agents, n8n flows, and practical automation developed during and after the Wayfair externship.
Developed a digital-twin workflow to predict temperature, dew point, and relative humidity from prior sensor states, cyclical time features, and physics-informed calculations. I then compared predicted and actual readings to generate residual features for detecting sensor faults.
Building intelligent applications with machine learning, LLMs, RAG, computer vision, and automation.
I'm an AI/ML and Software Engineer with an M.S. in Computer Science and Software Engineering from Auburn University. I enjoy building intelligent applications that combine machine learning, LLMs, RAG, computer vision, automation, and modern software development. My work ranges from AI-powered applications and time-series anomaly detection to test automation and full-stack projects. I'm especially interested in turning AI concepts into practical systems that solve real problems, while continuously experimenting with new tools, models, and technologies.
Externships
Wayfair n8n AI Agent Engineering Externship
Wayfair
Experience
Java Graduate Teaching Assistant
Auburn University · Jul 2024 - Dec 2024
Education
Auburn University
M.S. Computer Science & Software Engineering · Class of 2024
Skills
Developed a digital-twin workflow to predict temperature, dew point, and relative humidity from prior sensor states, cyclical time features, and physics-informed calculations. I then compared predicted and actual readings to generate residual features for detecting sensor faults.
Developed a digital-twin workflow to predict temperature, dew point, and relative humidity from prior sensor states, cyclical time features, and physics-informed calculations. I then compared predicted and actual readings to generate residual features for detecting sensor faults.
Engineered residual-based features and trained XGBoost, Logistic Regression, and Naive Bayes classifiers to identify spike, drift, stuck-at, and replay faults. I also evaluated Transformer-based approaches for improving real-time fault classification.
The XGBoost baseline achieved 0.88 accuracy and 0.99 anomaly precision, demonstrating that digital-twin predictions combined with ML can provide a strong foundation for real-time sensor monitoring and automated fault detection.