I'm a computer vision leader building models that see — from industrial inspection to plant phenotyping.
Find Out MoreI'm Director of Computer Vision at InFlightAI, where I lead a team taking computer vision models from research to production for industrial inspection. I also run Oxbow Solutions, helping agricultural research labs use data science and computer vision to speed up discovery and cut manual labor. Before that I spent a decade in plant science, most recently leading deep learning phenotyping at The Land Institute, so I'm equally at home with field trials, image pipelines, and deployment infrastructure.
See my top skills
My primary language for computer vision, data pipelines, and backend services
Detection and segmentation models, from semi-supervised training to production
Fast, reproducible experimentation and multi-GPU training
Image preprocessing, classical vision, and measurement pipelines
Classical ML, clustering, and dimensionality reduction
Mixed effects models, meta-analysis, and experimental design
Data wrangling and publication-quality visualization with ggplot2
Designing and running research databases behind deployed apps
Full-stack web applications for image upload, inference, and results
REST APIs for model inference and database access
Lightweight, high-performance model serving endpoints
Containerized training and inference environments
S3, EC2, and SageMaker for storage, compute, and deployment
Hosting web apps and serverless inference endpoints
Serverless GPU inference and on-demand training compute
Version control, code review, and open-source collaboration
Lead end-to-end development and deployment of computer vision models, from research through production, and set the technical vision and strategy for computer vision initiatives. Lead and mentor the Computer Vision Engineering team and interns, building scalable engineering practices and turning computer vision capabilities into products.
Built and trained computer vision models for industrial inspection, from data and research through production deployment.
Help agricultural research labs use data science and computer vision to accelerate discovery and reduce manual labor, designing and deploying machine learning solutions that turn large-scale field and imaging data into actionable insights.
Lead applications developer in the perennial legumes program. Led all deep learning plant phenotyping work (object detection, semantic segmentation, semi-supervised learning) and built serverless inference endpoints on AWS, DigitalOcean, and RunPod, along with full-stack web apps and REST APIs for model inference. Published open datasets, drove crop ontology and FAIR data practices, and mentored interns and annotation teams.
Managed a citrus plant physiology lab focused on rootstock research, root system architecture, and horticultural performance. Coordinated multi-year field, greenhouse, and lab studies, ran statistical modeling and image analysis in R and Python, managed grant-funded accounts, and supervised and trained staff and graduate students.
Evaluated disarmed citrus tristeza virus vectors in citrus trees, running qPCR gene expression assays, HPLC organic acid analysis, and greenhouse research.
Tomato grafting field trials across ~150 Solanaceae rootstocks, qPCR and marker work on citrus, and curation of a 400+ genotype citrus variety collection.