AI/ML Engineer working across LLMs & agents, machine learning, and computer vision — from RAG copilots serving 40,000+ users to an NSF-recognized clinical-AI system. Every model ships behind a real evaluation.
Four projects across healthcare AI, generative computer vision, and applied ML — each with a public repo or paper and metrics I can defend.
A multi-method clinical-NLP pipeline over 52,184 patient records that flags documentation and coding gaps, then traces them to missed ICD-10 codes and DRG downgrades — $27.5M–$60M in recoverable revenue at a mid-sized hospital.
An AI chatbot that answers drug-safety questions without hallucinating. A 4-tier agent (structured lookup → keyword → Random-Forest classifier → LLM+RAG fallback) escalates only when it must — reliable and cheap to run.
A KinematicEncoder maps 76-dim motion/kinematic features into a LoRA-fine-tuned Stable Diffusion U-Net (~1% trainable params) to generate gesture-accurate surgical video from the JIGSAWS robotics dataset.
Benchmarked CNN, QCNN, SVM, and QSVM on chest X-ray classification, with SHAP and Grad-CAM validating every prediction for clinician trust and failure analysis.
Everything listed here shows up in a project or a shipped system above — no résumé padding.
I'm an AI/ML engineer who likes the unglamorous half of the job — the evaluation, the deployment, the monitoring — as much as the modeling.
Over the past year at TIFIN, I took machine-learning and LLM systems from feature engineering all the way to production, where real users and real data broke them in ways no notebook predicts. My research and side projects run on the same principle: a model only counts once there's a real evaluation behind it. That's how ClinIQ went from an F1 score to a $27.5M–$60M revenue-cycle finding, and how my drug-interaction agent went from a demo to a deployed app answering in under 100 ms.
My work spans LLMs and agents, applied ML, and computer vision — most of it in healthcare, where the difference between "the model is accurate" and "here's what it changes for a patient or a hospital" actually matters. I completed my M.S. in Computer Science (AI emphasis) at UMKC in May 2026, and I'm looking for a team building things that reach real people.
I'm open to AI/ML Engineer roles and always happy to talk shop about agents, evaluation, or getting models into production.