I build AI systems from research through production. I have a PhD in Geophysics and eight years of industry experience across generative AI, speech systems, computer vision, medical imaging, and model optimization.
Selected results:
• Sirona Medical: reduced medical-image segmentation inference from 2–3 minutes to 4–10 seconds using ONNX and model optimization.
• Bridgestone: built tire X-ray anomaly detection delivering more than $20M in annual savings.
• Asurion: built an automated fraud-detection framework associated with $18M in annual savings.
• Toyota Connected: deployed an optimized small language model on constrained automotive hardware.
As founder of Ontor, I built the product and research workflow from the ground up, including on-device acoustic analysis, compact speech models, and a local ASR–small-language-model–TTS prototype. My work includes distillation, quantization, speech synthesis experiments, and evaluation across speakers and datasets.
I have led a team of five engineers and data scientists and taught undergraduate applied machine learning for one semester at Fisk University. I bring hands-on Python, PyTorch, ONNX, RAG, and production model-development experience.
Seeking Machine Learning Scientist, Machine Learning Engineer, or Applied AI roles. Open to full-time, contract, and part-time opportunities, including remote work. Based in Melissa, Texas (DFW), in Central Time.
Reduced medical-image segmentation inference from 2–3 minutes to 4–10 seconds using ONNX and model optimization. Developed CT/MRI registration using anatomical metadata, rigid alignment, and thin-plate-spline deformation. Built segmentation pipelines and contributed to medical language-model and multi-agent radiology workflows.
Deployed an optimized small language model on an automotive head unit under memory and latency constraints. Applied compression, quantization, and architecture tuning, and developed speech and audio components for in-vehicle conversational systems.
Led five engineers and data scientists developing sensor-based machine learning and AWS deployment infrastructure for an elderly-care platform. Developed fall-detection models from sensor time series and coordinated model development, evaluation, and delivery.
Developed a Medicare recommendation system and voice/text enrollment workflows using language models, retrieval-augmented generation (RAG), and multi-agent architectures. Connected model adaptation and information retrieval with application-specific evaluation and production delivery.
Built and deployed CNN/YOLO-based tire X-ray anomaly detection for manufacturing inspection, delivering more than $20M in annual savings. Conducted applied R&D in computer vision, reinforcement learning, NLP, and diffusion models.
Built and deployed an automated fraud-detection framework associated with $18M in annual savings. Applied graph methods, Bayesian deep learning, NLP, and tree-based models to fraud-pattern detection and high-volume data analysis.
Ph.D., Geophysics
Research in computational geophysics, numerical simulation, and machine learning for earthquake rupture. Dissertation: Computational Journey Across Multiphysics and Multitemporal Scales: From Slow Tectonic Movement to Fast Earthquake Slip.
M.Sc., Geological Sciences
Degree awarded July 2011.
B.Sc., Geological Sciences
Degree awarded July 2010.
This professional hasn’t added portfolio projects yet.
This professional hasn’t listed any services yet.
