At ZeroFox, I deploy machine learning models within a microservices architecture, including shadow deployments and health and performance monitoring. I also fine-tune Transformer models with LoRA and supervised fine-tuning for classification and semantic matching.
At NotCo, I worked with software engineers to develop and deploy cloud-based LLM services, including tool-agent chatbots and metadata extraction. I also designed, trained, and deployed two Transformer models for formula generation; the GAT model is applying for a patent for aroma development.
My master’s thesis proposed a convolution-based MultiHeadAttention mechanism for processing long time series, and I was a main contributor to a patent application for Sensory Transformer, an architecture for generating formulas from natural-language descriptions. I’ve also mentored junior machine learning engineers and guided an intern working on a multimodal compound representation project.
